Invalid JSON: Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
| { | |
| "best_global_step": 79000, | |
| "best_metric": 3.528618335723877, | |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/swap_require_to_hit_2128/checkpoint-40000", | |
| "epoch": 28.812761993479274, | |
| "eval_steps": 1000, | |
| "global_step": 99000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014555193292966931, | |
| "grad_norm": 0.9460402727127075, | |
| "learning_rate": 0.000294, | |
| "loss": 8.4453, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029110386585933862, | |
| "grad_norm": 0.7167385220527649, | |
| "learning_rate": 0.0005939999999999999, | |
| "loss": 6.7053, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04366557987890079, | |
| "grad_norm": 0.4383399784564972, | |
| "learning_rate": 0.0005998287711124053, | |
| "loss": 6.3193, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.058220773171867725, | |
| "grad_norm": 0.592664897441864, | |
| "learning_rate": 0.000599654047757717, | |
| "loss": 6.1411, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07277596646483465, | |
| "grad_norm": 0.5240400433540344, | |
| "learning_rate": 0.0005994793244030285, | |
| "loss": 5.9784, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.08733115975780158, | |
| "grad_norm": 0.5000967383384705, | |
| "learning_rate": 0.00059930460104834, | |
| "loss": 5.8537, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.10188635305076851, | |
| "grad_norm": 0.4281861484050751, | |
| "learning_rate": 0.0005991298776936517, | |
| "loss": 5.7267, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.11644154634373545, | |
| "grad_norm": 0.5191864967346191, | |
| "learning_rate": 0.0005989551543389632, | |
| "loss": 5.6177, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1309967396367024, | |
| "grad_norm": 0.42221343517303467, | |
| "learning_rate": 0.0005987804309842748, | |
| "loss": 5.4865, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1455519329296693, | |
| "grad_norm": 0.5270488262176514, | |
| "learning_rate": 0.0005986057076295864, | |
| "loss": 5.4108, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16010712622263623, | |
| "grad_norm": 0.39462053775787354, | |
| "learning_rate": 0.0005984309842748981, | |
| "loss": 5.3158, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.17466231951560315, | |
| "grad_norm": 0.4455013871192932, | |
| "learning_rate": 0.0005982562609202096, | |
| "loss": 5.2353, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1892175128085701, | |
| "grad_norm": 0.47907572984695435, | |
| "learning_rate": 0.0005980815375655212, | |
| "loss": 5.1769, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.20377270610153703, | |
| "grad_norm": 0.4356594979763031, | |
| "learning_rate": 0.0005979068142108328, | |
| "loss": 5.1334, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.21832789939450395, | |
| "grad_norm": 0.44511595368385315, | |
| "learning_rate": 0.0005977320908561445, | |
| "loss": 5.0796, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2328830926874709, | |
| "grad_norm": 0.5012144446372986, | |
| "learning_rate": 0.000597557367501456, | |
| "loss": 5.038, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.24743828598043782, | |
| "grad_norm": 0.3792232573032379, | |
| "learning_rate": 0.0005973826441467675, | |
| "loss": 4.9598, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2619934792734048, | |
| "grad_norm": 0.4020742177963257, | |
| "learning_rate": 0.0005972079207920792, | |
| "loss": 4.922, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.27654867256637167, | |
| "grad_norm": 0.44200628995895386, | |
| "learning_rate": 0.0005970331974373907, | |
| "loss": 4.8732, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "grad_norm": 0.45107322931289673, | |
| "learning_rate": 0.0005968584740827023, | |
| "loss": 4.8322, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "eval_accuracy": 0.2531876637228819, | |
| "eval_loss": 4.760279178619385, | |
| "eval_runtime": 82.7383, | |
| "eval_samples_per_second": 201.249, | |
| "eval_steps_per_second": 12.582, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.30565905915230557, | |
| "grad_norm": 0.5260637402534485, | |
| "learning_rate": 0.0005966837507280139, | |
| "loss": 4.7778, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.32021425244527246, | |
| "grad_norm": 0.43638306856155396, | |
| "learning_rate": 0.0005965090273733256, | |
| "loss": 4.7386, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3347694457382394, | |
| "grad_norm": 0.42931240797042847, | |
| "learning_rate": 0.0005963343040186371, | |
| "loss": 4.7026, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3493246390312063, | |
| "grad_norm": 0.42878690361976624, | |
| "learning_rate": 0.0005961595806639486, | |
| "loss": 4.6624, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.36387983232417326, | |
| "grad_norm": 0.49616560339927673, | |
| "learning_rate": 0.0005959848573092603, | |
| "loss": 4.6309, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.3784350256171402, | |
| "grad_norm": 0.4818870723247528, | |
| "learning_rate": 0.0005958101339545718, | |
| "loss": 4.6121, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3929902189101071, | |
| "grad_norm": 0.5048239827156067, | |
| "learning_rate": 0.0005956354105998835, | |
| "loss": 4.5887, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.40754541220307405, | |
| "grad_norm": 0.4268573522567749, | |
| "learning_rate": 0.000595460687245195, | |
| "loss": 4.5622, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.422100605496041, | |
| "grad_norm": 0.4221375286579132, | |
| "learning_rate": 0.0005952859638905067, | |
| "loss": 4.5201, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.4366557987890079, | |
| "grad_norm": 0.4160709083080292, | |
| "learning_rate": 0.0005951112405358182, | |
| "loss": 4.4944, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.45121099208197485, | |
| "grad_norm": 0.3999669551849365, | |
| "learning_rate": 0.0005949365171811299, | |
| "loss": 4.4924, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4657661853749418, | |
| "grad_norm": 0.36177924275398254, | |
| "learning_rate": 0.0005947617938264414, | |
| "loss": 4.4545, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.4803213786679087, | |
| "grad_norm": 0.40998929738998413, | |
| "learning_rate": 0.000594587070471753, | |
| "loss": 4.443, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.49487657196087564, | |
| "grad_norm": 0.40750619769096375, | |
| "learning_rate": 0.0005944123471170646, | |
| "loss": 4.4227, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5094317652538426, | |
| "grad_norm": 0.43662741780281067, | |
| "learning_rate": 0.0005942376237623762, | |
| "loss": 4.4189, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5239869585468095, | |
| "grad_norm": 0.378104031085968, | |
| "learning_rate": 0.0005940629004076878, | |
| "loss": 4.3892, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5385421518397764, | |
| "grad_norm": 0.40325984358787537, | |
| "learning_rate": 0.0005938881770529993, | |
| "loss": 4.3837, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5530973451327433, | |
| "grad_norm": 0.3893411457538605, | |
| "learning_rate": 0.000593713453698311, | |
| "loss": 4.3769, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.5676525384257103, | |
| "grad_norm": 0.4048912525177002, | |
| "learning_rate": 0.0005935387303436226, | |
| "loss": 4.3429, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "grad_norm": 0.3891216218471527, | |
| "learning_rate": 0.0005933640069889342, | |
| "loss": 4.3302, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "eval_accuracy": 0.299134376137444, | |
| "eval_loss": 4.285494327545166, | |
| "eval_runtime": 82.287, | |
| "eval_samples_per_second": 202.353, | |
| "eval_steps_per_second": 12.651, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5967629250116442, | |
| "grad_norm": 0.4288868010044098, | |
| "learning_rate": 0.0005931892836342457, | |
| "loss": 4.3235, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.6113181183046111, | |
| "grad_norm": 0.3963388204574585, | |
| "learning_rate": 0.0005930145602795573, | |
| "loss": 4.3103, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.625873311597578, | |
| "grad_norm": 0.38704973459243774, | |
| "learning_rate": 0.000592839836924869, | |
| "loss": 4.2958, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6404285048905449, | |
| "grad_norm": 0.3845170736312866, | |
| "learning_rate": 0.0005926651135701805, | |
| "loss": 4.2876, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.6549836981835119, | |
| "grad_norm": 0.36367514729499817, | |
| "learning_rate": 0.0005924903902154921, | |
| "loss": 4.2685, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6695388914764788, | |
| "grad_norm": 0.3983514606952667, | |
| "learning_rate": 0.0005923156668608037, | |
| "loss": 4.2607, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.6840940847694458, | |
| "grad_norm": 0.3855116367340088, | |
| "learning_rate": 0.0005921409435061153, | |
| "loss": 4.2721, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6986492780624126, | |
| "grad_norm": 0.37813800573349, | |
| "learning_rate": 0.0005919662201514268, | |
| "loss": 4.2537, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7132044713553796, | |
| "grad_norm": 0.36541375517845154, | |
| "learning_rate": 0.0005917914967967384, | |
| "loss": 4.2253, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.7277596646483465, | |
| "grad_norm": 0.4002459943294525, | |
| "learning_rate": 0.0005916167734420501, | |
| "loss": 4.2316, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7423148579413135, | |
| "grad_norm": 0.369521826505661, | |
| "learning_rate": 0.0005914420500873616, | |
| "loss": 4.2049, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7568700512342804, | |
| "grad_norm": 0.3709098696708679, | |
| "learning_rate": 0.0005912673267326732, | |
| "loss": 4.2106, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.7714252445272474, | |
| "grad_norm": 0.3567032516002655, | |
| "learning_rate": 0.0005910926033779848, | |
| "loss": 4.2007, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.7859804378202142, | |
| "grad_norm": 0.38760271668434143, | |
| "learning_rate": 0.0005909178800232964, | |
| "loss": 4.1826, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8005356311131812, | |
| "grad_norm": 0.3883505165576935, | |
| "learning_rate": 0.000590743156668608, | |
| "loss": 4.1806, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.8150908244061481, | |
| "grad_norm": 0.33989930152893066, | |
| "learning_rate": 0.0005905684333139196, | |
| "loss": 4.1624, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8296460176991151, | |
| "grad_norm": 0.35496529936790466, | |
| "learning_rate": 0.0005903937099592312, | |
| "loss": 4.1762, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.844201210992082, | |
| "grad_norm": 0.38042905926704407, | |
| "learning_rate": 0.0005902189866045427, | |
| "loss": 4.1587, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.858756404285049, | |
| "grad_norm": 0.347866028547287, | |
| "learning_rate": 0.0005900442632498543, | |
| "loss": 4.1604, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "grad_norm": 0.3856357932090759, | |
| "learning_rate": 0.0005898695398951659, | |
| "loss": 4.1451, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "eval_accuracy": 0.3155350765531733, | |
| "eval_loss": 4.095544338226318, | |
| "eval_runtime": 82.4036, | |
| "eval_samples_per_second": 202.066, | |
| "eval_steps_per_second": 12.633, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8878667908709827, | |
| "grad_norm": 0.3623169958591461, | |
| "learning_rate": 0.0005896948165404776, | |
| "loss": 4.1609, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.9024219841639497, | |
| "grad_norm": 0.3774421215057373, | |
| "learning_rate": 0.0005895200931857891, | |
| "loss": 4.1295, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9169771774569166, | |
| "grad_norm": 0.36384832859039307, | |
| "learning_rate": 0.0005893453698311007, | |
| "loss": 4.1356, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9315323707498836, | |
| "grad_norm": 0.3566246032714844, | |
| "learning_rate": 0.0005891706464764123, | |
| "loss": 4.1132, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.9460875640428504, | |
| "grad_norm": 0.354792058467865, | |
| "learning_rate": 0.0005889959231217238, | |
| "loss": 4.1122, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.9606427573358174, | |
| "grad_norm": 0.3386909067630768, | |
| "learning_rate": 0.0005888211997670355, | |
| "loss": 4.0995, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.9751979506287843, | |
| "grad_norm": 0.3349902629852295, | |
| "learning_rate": 0.000588646476412347, | |
| "loss": 4.087, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.9897531439217513, | |
| "grad_norm": 0.3581283688545227, | |
| "learning_rate": 0.0005884717530576587, | |
| "loss": 4.1023, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.0040754541220307, | |
| "grad_norm": 0.369205117225647, | |
| "learning_rate": 0.0005882970297029702, | |
| "loss": 4.0651, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.0186306474149978, | |
| "grad_norm": 0.36624059081077576, | |
| "learning_rate": 0.0005881223063482818, | |
| "loss": 4.0257, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.0331858407079646, | |
| "grad_norm": 0.3731226325035095, | |
| "learning_rate": 0.0005879475829935934, | |
| "loss": 4.0175, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.0477410340009314, | |
| "grad_norm": 0.33916571736335754, | |
| "learning_rate": 0.0005877728596389051, | |
| "loss": 4.0033, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.0622962272938985, | |
| "grad_norm": 0.3798449635505676, | |
| "learning_rate": 0.0005875981362842166, | |
| "loss": 4.0204, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.0768514205868653, | |
| "grad_norm": 0.3602329194545746, | |
| "learning_rate": 0.0005874234129295281, | |
| "loss": 4.0143, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.0914066138798324, | |
| "grad_norm": 0.3368425965309143, | |
| "learning_rate": 0.0005872486895748398, | |
| "loss": 4.0075, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.1059618071727992, | |
| "grad_norm": 0.3401786983013153, | |
| "learning_rate": 0.0005870739662201513, | |
| "loss": 3.9988, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.120517000465766, | |
| "grad_norm": 0.3401827812194824, | |
| "learning_rate": 0.000586899242865463, | |
| "loss": 3.9989, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.1350721937587331, | |
| "grad_norm": 0.3413340449333191, | |
| "learning_rate": 0.0005867245195107746, | |
| "loss": 4.0158, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.1496273870517, | |
| "grad_norm": 0.3464316427707672, | |
| "learning_rate": 0.0005865497961560862, | |
| "loss": 3.9982, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "grad_norm": 0.37090274691581726, | |
| "learning_rate": 0.0005863750728013977, | |
| "loss": 3.9892, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "eval_accuracy": 0.32568873057699677, | |
| "eval_loss": 3.9890823364257812, | |
| "eval_runtime": 82.3415, | |
| "eval_samples_per_second": 202.219, | |
| "eval_steps_per_second": 12.642, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.1787377736376339, | |
| "grad_norm": 0.3435836434364319, | |
| "learning_rate": 0.0005862003494467094, | |
| "loss": 3.9915, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.193292966930601, | |
| "grad_norm": 0.3374698758125305, | |
| "learning_rate": 0.0005860256260920209, | |
| "loss": 3.9741, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.2078481602235678, | |
| "grad_norm": 0.3420063257217407, | |
| "learning_rate": 0.0005858509027373325, | |
| "loss": 3.9723, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 1.2224033535165346, | |
| "grad_norm": 0.3533327579498291, | |
| "learning_rate": 0.0005856761793826441, | |
| "loss": 3.9784, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.2369585468095017, | |
| "grad_norm": 0.3362007737159729, | |
| "learning_rate": 0.0005855014560279557, | |
| "loss": 3.9678, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 1.2515137401024685, | |
| "grad_norm": 0.342079222202301, | |
| "learning_rate": 0.0005853267326732673, | |
| "loss": 3.9681, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.2660689333954354, | |
| "grad_norm": 0.3487435579299927, | |
| "learning_rate": 0.0005851520093185788, | |
| "loss": 3.9663, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 1.2806241266884024, | |
| "grad_norm": 0.3432266116142273, | |
| "learning_rate": 0.0005849772859638905, | |
| "loss": 3.9548, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.2951793199813695, | |
| "grad_norm": 0.33003994822502136, | |
| "learning_rate": 0.0005848025626092021, | |
| "loss": 3.9528, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 1.3097345132743363, | |
| "grad_norm": 0.3521156311035156, | |
| "learning_rate": 0.0005846278392545136, | |
| "loss": 3.9667, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.3242897065673032, | |
| "grad_norm": 0.3323107957839966, | |
| "learning_rate": 0.0005844531158998252, | |
| "loss": 3.9605, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.3388448998602702, | |
| "grad_norm": 0.3607315421104431, | |
| "learning_rate": 0.0005842783925451368, | |
| "loss": 3.9455, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.353400093153237, | |
| "grad_norm": 0.342742383480072, | |
| "learning_rate": 0.0005841036691904484, | |
| "loss": 3.9497, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.367955286446204, | |
| "grad_norm": 0.31976038217544556, | |
| "learning_rate": 0.00058392894583576, | |
| "loss": 3.9524, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.382510479739171, | |
| "grad_norm": 0.3248779773712158, | |
| "learning_rate": 0.0005837542224810716, | |
| "loss": 3.9521, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 1.3970656730321378, | |
| "grad_norm": 0.3700917959213257, | |
| "learning_rate": 0.0005835794991263832, | |
| "loss": 3.9424, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.4116208663251049, | |
| "grad_norm": 0.33306390047073364, | |
| "learning_rate": 0.0005834047757716948, | |
| "loss": 3.9385, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 1.4261760596180717, | |
| "grad_norm": 0.34592321515083313, | |
| "learning_rate": 0.0005832300524170063, | |
| "loss": 3.9475, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.4407312529110388, | |
| "grad_norm": 0.34128281474113464, | |
| "learning_rate": 0.0005830553290623179, | |
| "loss": 3.9221, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "grad_norm": 0.37158551812171936, | |
| "learning_rate": 0.0005828806057076296, | |
| "loss": 3.9311, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "eval_accuracy": 0.33233243162467047, | |
| "eval_loss": 3.9144339561462402, | |
| "eval_runtime": 82.3776, | |
| "eval_samples_per_second": 202.13, | |
| "eval_steps_per_second": 12.637, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4698416394969724, | |
| "grad_norm": 0.337547242641449, | |
| "learning_rate": 0.0005827058823529411, | |
| "loss": 3.9243, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 1.4843968327899395, | |
| "grad_norm": 0.34459587931632996, | |
| "learning_rate": 0.0005825311589982527, | |
| "loss": 3.9172, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.4989520260829063, | |
| "grad_norm": 0.3513347804546356, | |
| "learning_rate": 0.0005823564356435643, | |
| "loss": 3.9132, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 1.5135072193758732, | |
| "grad_norm": 0.32955098152160645, | |
| "learning_rate": 0.0005821817122888759, | |
| "loss": 3.9254, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.5280624126688402, | |
| "grad_norm": 0.33544817566871643, | |
| "learning_rate": 0.0005820069889341875, | |
| "loss": 3.9148, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 1.5426176059618073, | |
| "grad_norm": 0.3628521263599396, | |
| "learning_rate": 0.000581832265579499, | |
| "loss": 3.917, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.5571727992547741, | |
| "grad_norm": 0.3412168025970459, | |
| "learning_rate": 0.0005816575422248107, | |
| "loss": 3.9251, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 1.571727992547741, | |
| "grad_norm": 0.33038851618766785, | |
| "learning_rate": 0.0005814828188701222, | |
| "loss": 3.9163, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.586283185840708, | |
| "grad_norm": 0.3739718794822693, | |
| "learning_rate": 0.0005813080955154338, | |
| "loss": 3.9074, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.6008383791336749, | |
| "grad_norm": 0.3286726772785187, | |
| "learning_rate": 0.0005811333721607454, | |
| "loss": 3.9081, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.6153935724266417, | |
| "grad_norm": 0.3578357696533203, | |
| "learning_rate": 0.0005809586488060571, | |
| "loss": 3.9028, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.6299487657196088, | |
| "grad_norm": 0.3363804817199707, | |
| "learning_rate": 0.0005807839254513686, | |
| "loss": 3.8958, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.6445039590125758, | |
| "grad_norm": 0.3265131115913391, | |
| "learning_rate": 0.0005806092020966802, | |
| "loss": 3.8979, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.6590591523055425, | |
| "grad_norm": 0.3202497661113739, | |
| "learning_rate": 0.0005804344787419918, | |
| "loss": 3.8914, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.6736143455985095, | |
| "grad_norm": 0.36306965351104736, | |
| "learning_rate": 0.0005802597553873033, | |
| "loss": 3.9111, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.6881695388914766, | |
| "grad_norm": 0.33307918906211853, | |
| "learning_rate": 0.000580085032032615, | |
| "loss": 3.8888, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.7027247321844434, | |
| "grad_norm": 0.3406160771846771, | |
| "learning_rate": 0.0005799103086779265, | |
| "loss": 3.886, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.7172799254774103, | |
| "grad_norm": 0.3052501678466797, | |
| "learning_rate": 0.0005797355853232382, | |
| "loss": 3.8759, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.7318351187703773, | |
| "grad_norm": 0.33447790145874023, | |
| "learning_rate": 0.0005795608619685497, | |
| "loss": 3.8896, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "grad_norm": 0.31750261783599854, | |
| "learning_rate": 0.0005793861386138614, | |
| "loss": 3.8803, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "eval_accuracy": 0.33724977408313717, | |
| "eval_loss": 3.8548927307128906, | |
| "eval_runtime": 82.3728, | |
| "eval_samples_per_second": 202.142, | |
| "eval_steps_per_second": 12.638, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.760945505356311, | |
| "grad_norm": 0.344804048538208, | |
| "learning_rate": 0.0005792114152591729, | |
| "loss": 3.868, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.775500698649278, | |
| "grad_norm": 0.32179394364356995, | |
| "learning_rate": 0.0005790366919044846, | |
| "loss": 3.8701, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.7900558919422451, | |
| "grad_norm": 0.3431376814842224, | |
| "learning_rate": 0.0005788619685497961, | |
| "loss": 3.8796, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.804611085235212, | |
| "grad_norm": 0.3050195276737213, | |
| "learning_rate": 0.0005786872451951077, | |
| "loss": 3.8725, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.8191662785281788, | |
| "grad_norm": 0.32537171244621277, | |
| "learning_rate": 0.0005785125218404193, | |
| "loss": 3.8752, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.8337214718211459, | |
| "grad_norm": 0.3304484784603119, | |
| "learning_rate": 0.0005783377984857308, | |
| "loss": 3.8775, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.8482766651141127, | |
| "grad_norm": 0.3145851194858551, | |
| "learning_rate": 0.0005781630751310425, | |
| "loss": 3.8687, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.8628318584070795, | |
| "grad_norm": 0.32750535011291504, | |
| "learning_rate": 0.0005779883517763541, | |
| "loss": 3.8568, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.8773870517000466, | |
| "grad_norm": 0.34714996814727783, | |
| "learning_rate": 0.0005778136284216657, | |
| "loss": 3.8541, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.8919422449930137, | |
| "grad_norm": 0.3225823938846588, | |
| "learning_rate": 0.0005776389050669772, | |
| "loss": 3.8698, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.9064974382859803, | |
| "grad_norm": 0.3254883885383606, | |
| "learning_rate": 0.0005774641817122889, | |
| "loss": 3.8549, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.9210526315789473, | |
| "grad_norm": 0.31104394793510437, | |
| "learning_rate": 0.0005772894583576004, | |
| "loss": 3.8477, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.9356078248719144, | |
| "grad_norm": 0.3224423825740814, | |
| "learning_rate": 0.000577114735002912, | |
| "loss": 3.8468, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.9501630181648812, | |
| "grad_norm": 0.3226989805698395, | |
| "learning_rate": 0.0005769400116482236, | |
| "loss": 3.841, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.964718211457848, | |
| "grad_norm": 0.3361695408821106, | |
| "learning_rate": 0.0005767652882935352, | |
| "loss": 3.8461, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.9792734047508151, | |
| "grad_norm": 0.33380287885665894, | |
| "learning_rate": 0.0005765905649388468, | |
| "loss": 3.8484, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.993828598043782, | |
| "grad_norm": 0.3231297731399536, | |
| "learning_rate": 0.0005764158415841583, | |
| "loss": 3.8525, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 2.0081509082440614, | |
| "grad_norm": 0.3623596429824829, | |
| "learning_rate": 0.00057624111822947, | |
| "loss": 3.7972, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.0227061015370285, | |
| "grad_norm": 0.34392085671424866, | |
| "learning_rate": 0.0005760663948747816, | |
| "loss": 3.7518, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "grad_norm": 0.3092724680900574, | |
| "learning_rate": 0.0005758916715200931, | |
| "loss": 3.7506, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "eval_accuracy": 0.34154492698674915, | |
| "eval_loss": 3.8140785694122314, | |
| "eval_runtime": 82.3428, | |
| "eval_samples_per_second": 202.216, | |
| "eval_steps_per_second": 12.642, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.051816488122962, | |
| "grad_norm": 0.32952553033828735, | |
| "learning_rate": 0.0005757169481654047, | |
| "loss": 3.7478, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 2.066371681415929, | |
| "grad_norm": 0.3168792426586151, | |
| "learning_rate": 0.0005755422248107163, | |
| "loss": 3.7388, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.0809268747088963, | |
| "grad_norm": 0.30754348635673523, | |
| "learning_rate": 0.0005753675014560279, | |
| "loss": 3.7485, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 2.095482068001863, | |
| "grad_norm": 0.3121272325515747, | |
| "learning_rate": 0.0005751927781013395, | |
| "loss": 3.7478, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.11003726129483, | |
| "grad_norm": 0.3162359893321991, | |
| "learning_rate": 0.0005750180547466511, | |
| "loss": 3.7525, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 2.124592454587797, | |
| "grad_norm": 0.33430325984954834, | |
| "learning_rate": 0.0005748433313919627, | |
| "loss": 3.7458, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.139147647880764, | |
| "grad_norm": 0.3306877017021179, | |
| "learning_rate": 0.0005746686080372743, | |
| "loss": 3.7534, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 2.1537028411737307, | |
| "grad_norm": 0.328163743019104, | |
| "learning_rate": 0.0005744938846825858, | |
| "loss": 3.7623, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.1682580344666977, | |
| "grad_norm": 0.34212759137153625, | |
| "learning_rate": 0.0005743191613278974, | |
| "loss": 3.7653, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 2.182813227759665, | |
| "grad_norm": 0.31965479254722595, | |
| "learning_rate": 0.0005741444379732091, | |
| "loss": 3.7525, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.1973684210526314, | |
| "grad_norm": 0.3307754695415497, | |
| "learning_rate": 0.0005739697146185206, | |
| "loss": 3.761, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 2.2119236143455985, | |
| "grad_norm": 0.3454848527908325, | |
| "learning_rate": 0.0005737949912638322, | |
| "loss": 3.7481, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.2264788076385655, | |
| "grad_norm": 0.32161134481430054, | |
| "learning_rate": 0.0005736202679091438, | |
| "loss": 3.7491, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 2.241034000931532, | |
| "grad_norm": 0.3180389702320099, | |
| "learning_rate": 0.0005734455445544554, | |
| "loss": 3.7455, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.255589194224499, | |
| "grad_norm": 0.32045218348503113, | |
| "learning_rate": 0.000573270821199767, | |
| "loss": 3.753, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 2.2701443875174663, | |
| "grad_norm": 0.3247104287147522, | |
| "learning_rate": 0.0005730960978450785, | |
| "loss": 3.7566, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.2846995808104333, | |
| "grad_norm": 0.3317783772945404, | |
| "learning_rate": 0.0005729213744903902, | |
| "loss": 3.7531, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 2.2992547741034, | |
| "grad_norm": 0.34790417551994324, | |
| "learning_rate": 0.0005727466511357017, | |
| "loss": 3.7589, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.313809967396367, | |
| "grad_norm": 0.3196486234664917, | |
| "learning_rate": 0.0005725719277810134, | |
| "loss": 3.751, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "grad_norm": 0.31746625900268555, | |
| "learning_rate": 0.0005723972044263249, | |
| "loss": 3.7626, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "eval_accuracy": 0.3445834779408887, | |
| "eval_loss": 3.7837178707122803, | |
| "eval_runtime": 82.3474, | |
| "eval_samples_per_second": 202.204, | |
| "eval_steps_per_second": 12.642, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.3429203539823007, | |
| "grad_norm": 0.3117084801197052, | |
| "learning_rate": 0.0005722224810716366, | |
| "loss": 3.7336, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 2.3574755472752678, | |
| "grad_norm": 0.31918084621429443, | |
| "learning_rate": 0.0005720477577169481, | |
| "loss": 3.754, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.372030740568235, | |
| "grad_norm": 0.32155799865722656, | |
| "learning_rate": 0.0005718730343622598, | |
| "loss": 3.7532, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 2.386585933861202, | |
| "grad_norm": 0.3052207827568054, | |
| "learning_rate": 0.0005716983110075713, | |
| "loss": 3.754, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.4011411271541685, | |
| "grad_norm": 0.3154080808162689, | |
| "learning_rate": 0.0005715235876528828, | |
| "loss": 3.7502, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 2.4156963204471356, | |
| "grad_norm": 0.31491243839263916, | |
| "learning_rate": 0.0005713488642981945, | |
| "loss": 3.7555, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.4302515137401026, | |
| "grad_norm": 0.33506786823272705, | |
| "learning_rate": 0.0005711741409435061, | |
| "loss": 3.7526, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 2.4448067070330692, | |
| "grad_norm": 0.3143409192562103, | |
| "learning_rate": 0.0005709994175888177, | |
| "loss": 3.7542, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.4593619003260363, | |
| "grad_norm": 0.32687249779701233, | |
| "learning_rate": 0.0005708246942341292, | |
| "loss": 3.7586, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 2.4739170936190034, | |
| "grad_norm": 0.313556969165802, | |
| "learning_rate": 0.0005706499708794409, | |
| "loss": 3.7619, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.4884722869119704, | |
| "grad_norm": 0.31588125228881836, | |
| "learning_rate": 0.0005704752475247524, | |
| "loss": 3.7356, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 2.503027480204937, | |
| "grad_norm": 0.32753488421440125, | |
| "learning_rate": 0.0005703005241700641, | |
| "loss": 3.7564, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.517582673497904, | |
| "grad_norm": 0.3250143826007843, | |
| "learning_rate": 0.0005701258008153756, | |
| "loss": 3.7452, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 2.5321378667908707, | |
| "grad_norm": 0.32976436614990234, | |
| "learning_rate": 0.0005699510774606872, | |
| "loss": 3.7289, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 2.546693060083838, | |
| "grad_norm": 0.3077791631221771, | |
| "learning_rate": 0.0005697763541059988, | |
| "loss": 3.741, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 2.561248253376805, | |
| "grad_norm": 0.32112279534339905, | |
| "learning_rate": 0.0005696016307513103, | |
| "loss": 3.7481, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.575803446669772, | |
| "grad_norm": 0.321278840303421, | |
| "learning_rate": 0.000569426907396622, | |
| "loss": 3.7297, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 2.590358639962739, | |
| "grad_norm": 0.32353463768959045, | |
| "learning_rate": 0.0005692521840419336, | |
| "loss": 3.7434, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 2.6049138332557056, | |
| "grad_norm": 0.313834011554718, | |
| "learning_rate": 0.0005690774606872452, | |
| "loss": 3.7363, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "grad_norm": 0.3230018615722656, | |
| "learning_rate": 0.0005689027373325567, | |
| "loss": 3.7455, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "eval_accuracy": 0.34739766926899546, | |
| "eval_loss": 3.7543318271636963, | |
| "eval_runtime": 82.369, | |
| "eval_samples_per_second": 202.151, | |
| "eval_steps_per_second": 12.638, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6340242198416393, | |
| "grad_norm": 0.3046814501285553, | |
| "learning_rate": 0.0005687280139778683, | |
| "loss": 3.7454, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 2.6485794131346063, | |
| "grad_norm": 0.32191336154937744, | |
| "learning_rate": 0.0005685532906231799, | |
| "loss": 3.7306, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 2.6631346064275734, | |
| "grad_norm": 0.32236990332603455, | |
| "learning_rate": 0.0005683785672684915, | |
| "loss": 3.7294, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 2.6776897997205404, | |
| "grad_norm": 0.3283775746822357, | |
| "learning_rate": 0.0005682038439138031, | |
| "loss": 3.7478, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.692244993013507, | |
| "grad_norm": 0.3124407231807709, | |
| "learning_rate": 0.0005680291205591147, | |
| "loss": 3.7498, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 2.706800186306474, | |
| "grad_norm": 0.325979620218277, | |
| "learning_rate": 0.0005678543972044263, | |
| "loss": 3.7423, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 2.721355379599441, | |
| "grad_norm": 0.3154834806919098, | |
| "learning_rate": 0.0005676796738497378, | |
| "loss": 3.7351, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 2.735910572892408, | |
| "grad_norm": 0.31207218766212463, | |
| "learning_rate": 0.0005675049504950495, | |
| "loss": 3.7291, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.750465766185375, | |
| "grad_norm": 0.336139440536499, | |
| "learning_rate": 0.0005673302271403611, | |
| "loss": 3.7414, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 2.765020959478342, | |
| "grad_norm": 0.30878597497940063, | |
| "learning_rate": 0.0005671555037856726, | |
| "loss": 3.7245, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.779576152771309, | |
| "grad_norm": 0.3130248486995697, | |
| "learning_rate": 0.0005669807804309842, | |
| "loss": 3.7305, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 2.7941313460642756, | |
| "grad_norm": 0.3261592984199524, | |
| "learning_rate": 0.0005668060570762958, | |
| "loss": 3.7325, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 2.8086865393572427, | |
| "grad_norm": 0.3152883052825928, | |
| "learning_rate": 0.0005666313337216074, | |
| "loss": 3.7256, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 2.8232417326502097, | |
| "grad_norm": 0.30999112129211426, | |
| "learning_rate": 0.000566456610366919, | |
| "loss": 3.7322, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 2.8377969259431763, | |
| "grad_norm": 0.3101558983325958, | |
| "learning_rate": 0.0005662818870122306, | |
| "loss": 3.7088, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 2.8523521192361434, | |
| "grad_norm": 0.3394653797149658, | |
| "learning_rate": 0.0005661071636575422, | |
| "loss": 3.7415, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 2.8669073125291105, | |
| "grad_norm": 0.3301781713962555, | |
| "learning_rate": 0.0005659324403028537, | |
| "loss": 3.7302, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 2.8814625058220775, | |
| "grad_norm": 0.3048422634601593, | |
| "learning_rate": 0.0005657577169481653, | |
| "loss": 3.718, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 2.896017699115044, | |
| "grad_norm": 0.3374245762825012, | |
| "learning_rate": 0.0005655829935934769, | |
| "loss": 3.7213, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "grad_norm": 0.30655840039253235, | |
| "learning_rate": 0.0005654082702387886, | |
| "loss": 3.7292, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "eval_accuracy": 0.34949529661600104, | |
| "eval_loss": 3.7286386489868164, | |
| "eval_runtime": 82.3005, | |
| "eval_samples_per_second": 202.319, | |
| "eval_steps_per_second": 12.649, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.9251280857009783, | |
| "grad_norm": 0.31123921275138855, | |
| "learning_rate": 0.0005652335468841001, | |
| "loss": 3.7165, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 2.939683278993945, | |
| "grad_norm": 0.32680487632751465, | |
| "learning_rate": 0.0005650588235294117, | |
| "loss": 3.7286, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 2.954238472286912, | |
| "grad_norm": 0.35948458313941956, | |
| "learning_rate": 0.0005648841001747233, | |
| "loss": 3.7138, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 2.968793665579879, | |
| "grad_norm": 0.32463008165359497, | |
| "learning_rate": 0.0005647093768200349, | |
| "loss": 3.7195, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 2.983348858872846, | |
| "grad_norm": 0.30369722843170166, | |
| "learning_rate": 0.0005645346534653465, | |
| "loss": 3.7079, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 2.9979040521658127, | |
| "grad_norm": 0.30367618799209595, | |
| "learning_rate": 0.0005643599301106582, | |
| "loss": 3.7114, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 3.0122263623660923, | |
| "grad_norm": 0.32326823472976685, | |
| "learning_rate": 0.0005641852067559697, | |
| "loss": 3.6301, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 3.026781555659059, | |
| "grad_norm": 0.33145368099212646, | |
| "learning_rate": 0.0005640104834012812, | |
| "loss": 3.6054, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 3.041336748952026, | |
| "grad_norm": 0.34357044100761414, | |
| "learning_rate": 0.0005638357600465929, | |
| "loss": 3.6166, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 3.055891942244993, | |
| "grad_norm": 0.34022730588912964, | |
| "learning_rate": 0.0005636610366919044, | |
| "loss": 3.6017, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.07044713553796, | |
| "grad_norm": 0.3136366009712219, | |
| "learning_rate": 0.0005634863133372161, | |
| "loss": 3.6096, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 3.0850023288309267, | |
| "grad_norm": 0.30901193618774414, | |
| "learning_rate": 0.0005633115899825276, | |
| "loss": 3.6285, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 3.099557522123894, | |
| "grad_norm": 0.3291122317314148, | |
| "learning_rate": 0.0005631368666278393, | |
| "loss": 3.6312, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 3.114112715416861, | |
| "grad_norm": 0.333393931388855, | |
| "learning_rate": 0.0005629621432731508, | |
| "loss": 3.6302, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 3.1286679087098275, | |
| "grad_norm": 0.3243166506290436, | |
| "learning_rate": 0.0005627874199184623, | |
| "loss": 3.6292, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 3.1432231020027945, | |
| "grad_norm": 0.3299334645271301, | |
| "learning_rate": 0.000562612696563774, | |
| "loss": 3.626, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 3.1577782952957616, | |
| "grad_norm": 0.33389437198638916, | |
| "learning_rate": 0.0005624379732090856, | |
| "loss": 3.6317, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 3.1723334885887287, | |
| "grad_norm": 0.34346115589141846, | |
| "learning_rate": 0.0005622632498543972, | |
| "loss": 3.6336, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 3.1868886818816953, | |
| "grad_norm": 0.31092044711112976, | |
| "learning_rate": 0.0005620885264997087, | |
| "loss": 3.626, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 3.2014438751746623, | |
| "grad_norm": 0.3373558223247528, | |
| "learning_rate": 0.0005619138031450204, | |
| "loss": 3.6292, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2014438751746623, | |
| "eval_accuracy": 0.35180176998472495, | |
| "eval_loss": 3.7138261795043945, | |
| "eval_runtime": 82.4364, | |
| "eval_samples_per_second": 201.986, | |
| "eval_steps_per_second": 12.628, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2159990684676294, | |
| "grad_norm": 0.3097820580005646, | |
| "learning_rate": 0.0005617390797903319, | |
| "loss": 3.6323, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 3.230554261760596, | |
| "grad_norm": 0.3223876953125, | |
| "learning_rate": 0.0005615643564356436, | |
| "loss": 3.6418, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 3.245109455053563, | |
| "grad_norm": 0.31886398792266846, | |
| "learning_rate": 0.0005613896330809551, | |
| "loss": 3.6482, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 3.25966464834653, | |
| "grad_norm": 0.33520999550819397, | |
| "learning_rate": 0.0005612149097262667, | |
| "loss": 3.6306, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 3.274219841639497, | |
| "grad_norm": 0.3203602433204651, | |
| "learning_rate": 0.0005610401863715783, | |
| "loss": 3.6491, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 3.288775034932464, | |
| "grad_norm": 0.30814459919929504, | |
| "learning_rate": 0.0005608654630168898, | |
| "loss": 3.6406, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 3.303330228225431, | |
| "grad_norm": 0.32829371094703674, | |
| "learning_rate": 0.0005606907396622015, | |
| "loss": 3.6412, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 3.317885421518398, | |
| "grad_norm": 0.3164595663547516, | |
| "learning_rate": 0.0005605160163075131, | |
| "loss": 3.6445, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 3.3324406148113646, | |
| "grad_norm": 0.33146291971206665, | |
| "learning_rate": 0.0005603412929528247, | |
| "loss": 3.6456, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 3.3469958081043316, | |
| "grad_norm": 0.3133467435836792, | |
| "learning_rate": 0.0005601665695981362, | |
| "loss": 3.6383, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.3615510013972987, | |
| "grad_norm": 0.32508066296577454, | |
| "learning_rate": 0.0005599918462434478, | |
| "loss": 3.6367, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 3.3761061946902653, | |
| "grad_norm": 0.3299286961555481, | |
| "learning_rate": 0.0005598171228887594, | |
| "loss": 3.6453, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 3.3906613879832324, | |
| "grad_norm": 0.3155434727668762, | |
| "learning_rate": 0.0005596423995340709, | |
| "loss": 3.6357, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 3.4052165812761994, | |
| "grad_norm": 0.31516334414482117, | |
| "learning_rate": 0.0005594676761793826, | |
| "loss": 3.6406, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 3.419771774569166, | |
| "grad_norm": 0.33438608050346375, | |
| "learning_rate": 0.0005592929528246942, | |
| "loss": 3.6562, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 3.434326967862133, | |
| "grad_norm": 0.31893497705459595, | |
| "learning_rate": 0.0005591182294700058, | |
| "loss": 3.6426, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 3.4488821611551, | |
| "grad_norm": 0.3268669843673706, | |
| "learning_rate": 0.0005589435061153173, | |
| "loss": 3.6417, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 3.463437354448067, | |
| "grad_norm": 0.31991738080978394, | |
| "learning_rate": 0.000558768782760629, | |
| "loss": 3.6509, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 3.477992547741034, | |
| "grad_norm": 0.30439630150794983, | |
| "learning_rate": 0.0005585940594059406, | |
| "loss": 3.6514, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 3.492547741034001, | |
| "grad_norm": 0.31228959560394287, | |
| "learning_rate": 0.0005584193360512521, | |
| "loss": 3.6496, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.492547741034001, | |
| "eval_accuracy": 0.3534779444145207, | |
| "eval_loss": 3.694697380065918, | |
| "eval_runtime": 82.4272, | |
| "eval_samples_per_second": 202.008, | |
| "eval_steps_per_second": 12.629, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.507102934326968, | |
| "grad_norm": 0.32010728120803833, | |
| "learning_rate": 0.0005582446126965637, | |
| "loss": 3.6434, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 3.5216581276199346, | |
| "grad_norm": 0.31888964772224426, | |
| "learning_rate": 0.0005580698893418753, | |
| "loss": 3.6483, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 3.5362133209129016, | |
| "grad_norm": 0.31620028614997864, | |
| "learning_rate": 0.0005578951659871869, | |
| "loss": 3.6412, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 3.5507685142058687, | |
| "grad_norm": 0.31652969121932983, | |
| "learning_rate": 0.0005577204426324985, | |
| "loss": 3.6497, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 3.5653237074988358, | |
| "grad_norm": 0.3181188404560089, | |
| "learning_rate": 0.0005575457192778101, | |
| "loss": 3.6347, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 3.5798789007918024, | |
| "grad_norm": 0.3139238655567169, | |
| "learning_rate": 0.0005573709959231217, | |
| "loss": 3.6466, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 3.5944340940847694, | |
| "grad_norm": 0.3042029142379761, | |
| "learning_rate": 0.0005571962725684332, | |
| "loss": 3.651, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 3.6089892873777365, | |
| "grad_norm": 0.33461815118789673, | |
| "learning_rate": 0.0005570215492137449, | |
| "loss": 3.6415, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 3.623544480670703, | |
| "grad_norm": 0.31976258754730225, | |
| "learning_rate": 0.0005568468258590564, | |
| "loss": 3.6468, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 3.63809967396367, | |
| "grad_norm": 0.3129153251647949, | |
| "learning_rate": 0.0005566721025043681, | |
| "loss": 3.6333, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.6526548672566372, | |
| "grad_norm": 0.31318944692611694, | |
| "learning_rate": 0.0005564973791496796, | |
| "loss": 3.6498, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 3.6672100605496043, | |
| "grad_norm": 0.3211837112903595, | |
| "learning_rate": 0.0005563226557949913, | |
| "loss": 3.634, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 3.681765253842571, | |
| "grad_norm": 0.3278801739215851, | |
| "learning_rate": 0.0005561479324403028, | |
| "loss": 3.6386, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 3.696320447135538, | |
| "grad_norm": 0.30646154284477234, | |
| "learning_rate": 0.0005559732090856144, | |
| "loss": 3.6271, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 3.710875640428505, | |
| "grad_norm": 0.32493856549263, | |
| "learning_rate": 0.000555798485730926, | |
| "loss": 3.6561, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 3.7254308337214717, | |
| "grad_norm": 0.3191986083984375, | |
| "learning_rate": 0.0005556237623762376, | |
| "loss": 3.6486, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 3.7399860270144387, | |
| "grad_norm": 0.3006956875324249, | |
| "learning_rate": 0.0005554490390215492, | |
| "loss": 3.6329, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 3.754541220307406, | |
| "grad_norm": 0.3157041072845459, | |
| "learning_rate": 0.0005552743156668607, | |
| "loss": 3.636, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 3.769096413600373, | |
| "grad_norm": 0.3180425763130188, | |
| "learning_rate": 0.0005550995923121724, | |
| "loss": 3.6333, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 3.7836516068933395, | |
| "grad_norm": 0.31799381971359253, | |
| "learning_rate": 0.0005549248689574839, | |
| "loss": 3.6347, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.7836516068933395, | |
| "eval_accuracy": 0.3547632230265138, | |
| "eval_loss": 3.680462121963501, | |
| "eval_runtime": 82.3401, | |
| "eval_samples_per_second": 202.222, | |
| "eval_steps_per_second": 12.643, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.7982068001863065, | |
| "grad_norm": 0.3071649670600891, | |
| "learning_rate": 0.0005547501456027955, | |
| "loss": 3.6463, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 3.812761993479273, | |
| "grad_norm": 0.3161468803882599, | |
| "learning_rate": 0.0005545754222481071, | |
| "loss": 3.6535, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 3.82731718677224, | |
| "grad_norm": 0.3012949824333191, | |
| "learning_rate": 0.0005544006988934188, | |
| "loss": 3.6385, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 3.8418723800652073, | |
| "grad_norm": 0.3110561668872833, | |
| "learning_rate": 0.0005542259755387303, | |
| "loss": 3.6464, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 3.8564275733581743, | |
| "grad_norm": 0.32216086983680725, | |
| "learning_rate": 0.0005540512521840418, | |
| "loss": 3.6439, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 3.8709827666511414, | |
| "grad_norm": 0.3167315423488617, | |
| "learning_rate": 0.0005538765288293535, | |
| "loss": 3.637, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 3.885537959944108, | |
| "grad_norm": 0.3268546462059021, | |
| "learning_rate": 0.0005537018054746651, | |
| "loss": 3.6413, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 3.900093153237075, | |
| "grad_norm": 0.30385085940361023, | |
| "learning_rate": 0.0005535270821199767, | |
| "loss": 3.6392, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 3.9146483465300417, | |
| "grad_norm": 0.30552470684051514, | |
| "learning_rate": 0.0005533523587652882, | |
| "loss": 3.6317, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 3.9292035398230087, | |
| "grad_norm": 0.29734671115875244, | |
| "learning_rate": 0.0005531776354105999, | |
| "loss": 3.638, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.943758733115976, | |
| "grad_norm": 0.3026958107948303, | |
| "learning_rate": 0.0005530029120559114, | |
| "loss": 3.6441, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 3.958313926408943, | |
| "grad_norm": 0.30160197615623474, | |
| "learning_rate": 0.0005528281887012229, | |
| "loss": 3.6422, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 3.9728691197019095, | |
| "grad_norm": 0.3105238378047943, | |
| "learning_rate": 0.0005526534653465346, | |
| "loss": 3.6393, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 3.9874243129948765, | |
| "grad_norm": 0.3029833436012268, | |
| "learning_rate": 0.0005524787419918462, | |
| "loss": 3.6356, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 4.001746623195156, | |
| "grad_norm": 0.31539982557296753, | |
| "learning_rate": 0.0005523040186371578, | |
| "loss": 3.6264, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 4.016301816488123, | |
| "grad_norm": 0.32765814661979675, | |
| "learning_rate": 0.0005521292952824693, | |
| "loss": 3.5205, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 4.03085700978109, | |
| "grad_norm": 0.3211386799812317, | |
| "learning_rate": 0.000551954571927781, | |
| "loss": 3.5281, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 4.045412203074057, | |
| "grad_norm": 0.3320051431655884, | |
| "learning_rate": 0.0005517798485730926, | |
| "loss": 3.522, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 4.059967396367024, | |
| "grad_norm": 0.3103773593902588, | |
| "learning_rate": 0.0005516051252184042, | |
| "loss": 3.5312, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 4.074522589659991, | |
| "grad_norm": 0.31167536973953247, | |
| "learning_rate": 0.0005514304018637157, | |
| "loss": 3.5335, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.074522589659991, | |
| "eval_accuracy": 0.35640554959235066, | |
| "eval_loss": 3.674363374710083, | |
| "eval_runtime": 82.3732, | |
| "eval_samples_per_second": 202.141, | |
| "eval_steps_per_second": 12.638, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.089077782952957, | |
| "grad_norm": 0.3175770938396454, | |
| "learning_rate": 0.0005512556785090273, | |
| "loss": 3.5388, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 4.103632976245924, | |
| "grad_norm": 0.3252900540828705, | |
| "learning_rate": 0.0005510809551543389, | |
| "loss": 3.5461, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 4.118188169538891, | |
| "grad_norm": 0.29477643966674805, | |
| "learning_rate": 0.0005509062317996504, | |
| "loss": 3.5499, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 4.132743362831858, | |
| "grad_norm": 0.3362492322921753, | |
| "learning_rate": 0.0005507315084449621, | |
| "loss": 3.5458, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 4.1472985561248255, | |
| "grad_norm": 0.3338944613933563, | |
| "learning_rate": 0.0005505567850902737, | |
| "loss": 3.5485, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 4.1618537494177925, | |
| "grad_norm": 0.3101586401462555, | |
| "learning_rate": 0.0005503820617355853, | |
| "loss": 3.5621, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 4.17640894271076, | |
| "grad_norm": 0.31442511081695557, | |
| "learning_rate": 0.0005502073383808969, | |
| "loss": 3.5652, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 4.190964136003726, | |
| "grad_norm": 0.3452907204627991, | |
| "learning_rate": 0.0005500326150262085, | |
| "loss": 3.5537, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 4.205519329296693, | |
| "grad_norm": 0.33091723918914795, | |
| "learning_rate": 0.00054985789167152, | |
| "loss": 3.5604, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 4.22007452258966, | |
| "grad_norm": 0.3162982761859894, | |
| "learning_rate": 0.0005496831683168316, | |
| "loss": 3.5723, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.234629715882627, | |
| "grad_norm": 0.3247874081134796, | |
| "learning_rate": 0.0005495084449621433, | |
| "loss": 3.554, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 4.249184909175594, | |
| "grad_norm": 0.31667330861091614, | |
| "learning_rate": 0.0005493337216074548, | |
| "loss": 3.5504, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 4.263740102468561, | |
| "grad_norm": 0.30584612488746643, | |
| "learning_rate": 0.0005491589982527664, | |
| "loss": 3.5629, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 4.278295295761528, | |
| "grad_norm": 0.31260180473327637, | |
| "learning_rate": 0.000548984274898078, | |
| "loss": 3.56, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 4.292850489054494, | |
| "grad_norm": 0.3377438187599182, | |
| "learning_rate": 0.0005488095515433897, | |
| "loss": 3.5596, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 4.307405682347461, | |
| "grad_norm": 0.3167860805988312, | |
| "learning_rate": 0.0005486348281887012, | |
| "loss": 3.5686, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 4.321960875640428, | |
| "grad_norm": 0.3225920796394348, | |
| "learning_rate": 0.0005484601048340127, | |
| "loss": 3.5633, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 4.3365160689333955, | |
| "grad_norm": 0.31976449489593506, | |
| "learning_rate": 0.0005482853814793244, | |
| "loss": 3.5649, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 4.3510712622263625, | |
| "grad_norm": 0.316983163356781, | |
| "learning_rate": 0.0005481106581246359, | |
| "loss": 3.5706, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 4.36562645551933, | |
| "grad_norm": 0.33317041397094727, | |
| "learning_rate": 0.0005479359347699475, | |
| "loss": 3.5688, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.36562645551933, | |
| "eval_accuracy": 0.35754603456407535, | |
| "eval_loss": 3.6594462394714355, | |
| "eval_runtime": 82.3358, | |
| "eval_samples_per_second": 202.233, | |
| "eval_steps_per_second": 12.643, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.380181648812297, | |
| "grad_norm": 0.3106541335582733, | |
| "learning_rate": 0.0005477612114152591, | |
| "loss": 3.5737, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 4.394736842105263, | |
| "grad_norm": 0.3131338953971863, | |
| "learning_rate": 0.0005475864880605708, | |
| "loss": 3.5727, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 4.40929203539823, | |
| "grad_norm": 0.3236541450023651, | |
| "learning_rate": 0.0005474117647058823, | |
| "loss": 3.5697, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 4.423847228691197, | |
| "grad_norm": 0.31134647130966187, | |
| "learning_rate": 0.0005472370413511939, | |
| "loss": 3.5715, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 4.438402421984164, | |
| "grad_norm": 0.30151379108428955, | |
| "learning_rate": 0.0005470623179965055, | |
| "loss": 3.5725, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 4.452957615277131, | |
| "grad_norm": 0.3093511462211609, | |
| "learning_rate": 0.0005468875946418171, | |
| "loss": 3.5725, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 4.467512808570098, | |
| "grad_norm": 0.31018930673599243, | |
| "learning_rate": 0.0005467128712871287, | |
| "loss": 3.5653, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 4.482068001863064, | |
| "grad_norm": 0.31006720662117004, | |
| "learning_rate": 0.0005465381479324402, | |
| "loss": 3.5661, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 4.496623195156031, | |
| "grad_norm": 0.33606401085853577, | |
| "learning_rate": 0.0005463634245777519, | |
| "loss": 3.5746, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 4.511178388448998, | |
| "grad_norm": 0.31208300590515137, | |
| "learning_rate": 0.0005461887012230634, | |
| "loss": 3.5824, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 4.5257335817419655, | |
| "grad_norm": 0.29574018716812134, | |
| "learning_rate": 0.000546013977868375, | |
| "loss": 3.5792, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 4.5402887750349326, | |
| "grad_norm": 0.31001242995262146, | |
| "learning_rate": 0.0005458392545136866, | |
| "loss": 3.5741, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 4.5548439683279, | |
| "grad_norm": 0.3136127293109894, | |
| "learning_rate": 0.0005456645311589983, | |
| "loss": 3.5658, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 4.569399161620867, | |
| "grad_norm": 0.32919469475746155, | |
| "learning_rate": 0.0005454898078043098, | |
| "loss": 3.5847, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 4.583954354913834, | |
| "grad_norm": 0.32085543870925903, | |
| "learning_rate": 0.0005453150844496213, | |
| "loss": 3.5776, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 4.5985095482068, | |
| "grad_norm": 0.31507885456085205, | |
| "learning_rate": 0.000545140361094933, | |
| "loss": 3.5833, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 4.613064741499767, | |
| "grad_norm": 0.2956090271472931, | |
| "learning_rate": 0.0005449656377402445, | |
| "loss": 3.5882, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 4.627619934792734, | |
| "grad_norm": 0.3651985824108124, | |
| "learning_rate": 0.0005447909143855562, | |
| "loss": 3.5627, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 4.642175128085701, | |
| "grad_norm": 0.30227744579315186, | |
| "learning_rate": 0.0005446161910308677, | |
| "loss": 3.5708, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 4.656730321378668, | |
| "grad_norm": 0.30214428901672363, | |
| "learning_rate": 0.0005444414676761794, | |
| "loss": 3.5792, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.656730321378668, | |
| "eval_accuracy": 0.35873940682323574, | |
| "eval_loss": 3.647667407989502, | |
| "eval_runtime": 82.4075, | |
| "eval_samples_per_second": 202.057, | |
| "eval_steps_per_second": 12.632, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.671285514671635, | |
| "grad_norm": 0.3244737386703491, | |
| "learning_rate": 0.0005442667443214909, | |
| "loss": 3.5834, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 4.685840707964601, | |
| "grad_norm": 0.3233400881290436, | |
| "learning_rate": 0.0005440920209668024, | |
| "loss": 3.5764, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 4.7003959012575685, | |
| "grad_norm": 0.32228413224220276, | |
| "learning_rate": 0.0005439172976121141, | |
| "loss": 3.585, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 4.7149510945505355, | |
| "grad_norm": 0.3116596043109894, | |
| "learning_rate": 0.0005437425742574257, | |
| "loss": 3.564, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 4.729506287843503, | |
| "grad_norm": 0.31239327788352966, | |
| "learning_rate": 0.0005435678509027373, | |
| "loss": 3.5862, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 4.74406148113647, | |
| "grad_norm": 0.31308579444885254, | |
| "learning_rate": 0.0005433931275480488, | |
| "loss": 3.5805, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 4.758616674429437, | |
| "grad_norm": 0.31474950909614563, | |
| "learning_rate": 0.0005432184041933605, | |
| "loss": 3.5801, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 4.773171867722404, | |
| "grad_norm": 0.33123213052749634, | |
| "learning_rate": 0.000543043680838672, | |
| "loss": 3.5814, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 4.78772706101537, | |
| "grad_norm": 0.3135192096233368, | |
| "learning_rate": 0.0005428689574839837, | |
| "loss": 3.5723, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 4.802282254308337, | |
| "grad_norm": 0.3317807912826538, | |
| "learning_rate": 0.0005426942341292952, | |
| "loss": 3.5852, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 4.816837447601304, | |
| "grad_norm": 0.30853071808815, | |
| "learning_rate": 0.0005425195107746068, | |
| "loss": 3.581, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 4.831392640894271, | |
| "grad_norm": 0.31449249386787415, | |
| "learning_rate": 0.0005423447874199184, | |
| "loss": 3.5848, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 4.845947834187238, | |
| "grad_norm": 0.3099161386489868, | |
| "learning_rate": 0.00054217006406523, | |
| "loss": 3.5704, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 4.860503027480205, | |
| "grad_norm": 0.31092557311058044, | |
| "learning_rate": 0.0005419953407105417, | |
| "loss": 3.5814, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 4.875058220773171, | |
| "grad_norm": 0.3187222480773926, | |
| "learning_rate": 0.0005418206173558532, | |
| "loss": 3.5851, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 4.8896134140661385, | |
| "grad_norm": 0.3162148594856262, | |
| "learning_rate": 0.0005416458940011648, | |
| "loss": 3.5858, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 4.9041686073591055, | |
| "grad_norm": 0.3312950134277344, | |
| "learning_rate": 0.0005414711706464764, | |
| "loss": 3.5783, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 4.918723800652073, | |
| "grad_norm": 0.3099071979522705, | |
| "learning_rate": 0.000541296447291788, | |
| "loss": 3.5844, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 4.93327899394504, | |
| "grad_norm": 0.2964518964290619, | |
| "learning_rate": 0.0005411217239370995, | |
| "loss": 3.5775, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 4.947834187238007, | |
| "grad_norm": 0.29184818267822266, | |
| "learning_rate": 0.0005409470005824111, | |
| "loss": 3.5638, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.947834187238007, | |
| "eval_accuracy": 0.3599713280385715, | |
| "eval_loss": 3.633223533630371, | |
| "eval_runtime": 82.3155, | |
| "eval_samples_per_second": 202.283, | |
| "eval_steps_per_second": 12.646, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.962389380530974, | |
| "grad_norm": 0.3293011784553528, | |
| "learning_rate": 0.0005407722772277228, | |
| "loss": 3.5657, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 4.976944573823941, | |
| "grad_norm": 0.30635106563568115, | |
| "learning_rate": 0.0005405975538730343, | |
| "loss": 3.5668, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 4.991499767116907, | |
| "grad_norm": 0.3161274194717407, | |
| "learning_rate": 0.0005404228305183459, | |
| "loss": 3.5758, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 5.005822077317187, | |
| "grad_norm": 0.3411978483200073, | |
| "learning_rate": 0.0005402481071636575, | |
| "loss": 3.5347, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 5.020377270610154, | |
| "grad_norm": 0.30731824040412903, | |
| "learning_rate": 0.0005400733838089692, | |
| "loss": 3.4656, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 5.034932463903121, | |
| "grad_norm": 0.3365843892097473, | |
| "learning_rate": 0.0005398986604542807, | |
| "loss": 3.4693, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 5.049487657196088, | |
| "grad_norm": 0.3209933638572693, | |
| "learning_rate": 0.0005397239370995922, | |
| "loss": 3.4759, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 5.064042850489055, | |
| "grad_norm": 0.31448137760162354, | |
| "learning_rate": 0.0005395492137449039, | |
| "loss": 3.4725, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 5.078598043782021, | |
| "grad_norm": 0.32786670327186584, | |
| "learning_rate": 0.0005393744903902154, | |
| "loss": 3.4838, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 5.093153237074988, | |
| "grad_norm": 0.3159395456314087, | |
| "learning_rate": 0.000539199767035527, | |
| "loss": 3.4811, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 5.107708430367955, | |
| "grad_norm": 0.31278151273727417, | |
| "learning_rate": 0.0005390250436808386, | |
| "loss": 3.4806, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 5.122263623660922, | |
| "grad_norm": 0.3105210065841675, | |
| "learning_rate": 0.0005388503203261503, | |
| "loss": 3.5026, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 5.136818816953889, | |
| "grad_norm": 0.3473800718784332, | |
| "learning_rate": 0.0005386755969714618, | |
| "loss": 3.4848, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 5.151374010246856, | |
| "grad_norm": 0.3212640583515167, | |
| "learning_rate": 0.0005385008736167733, | |
| "loss": 3.4861, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 5.165929203539823, | |
| "grad_norm": 0.3321596086025238, | |
| "learning_rate": 0.000538326150262085, | |
| "loss": 3.4982, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 5.18048439683279, | |
| "grad_norm": 0.3351554274559021, | |
| "learning_rate": 0.0005381514269073965, | |
| "loss": 3.4739, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 5.195039590125757, | |
| "grad_norm": 0.3193203806877136, | |
| "learning_rate": 0.0005379767035527082, | |
| "loss": 3.4967, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 5.209594783418724, | |
| "grad_norm": 0.3100518584251404, | |
| "learning_rate": 0.0005378019801980197, | |
| "loss": 3.5099, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 5.224149976711691, | |
| "grad_norm": 0.3142613172531128, | |
| "learning_rate": 0.0005376272568433314, | |
| "loss": 3.4926, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 5.238705170004658, | |
| "grad_norm": 0.3395700752735138, | |
| "learning_rate": 0.0005374525334886429, | |
| "loss": 3.5018, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.238705170004658, | |
| "eval_accuracy": 0.36021895807107607, | |
| "eval_loss": 3.6357483863830566, | |
| "eval_runtime": 82.2873, | |
| "eval_samples_per_second": 202.352, | |
| "eval_steps_per_second": 12.651, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.253260363297625, | |
| "grad_norm": 0.3129439949989319, | |
| "learning_rate": 0.0005372778101339545, | |
| "loss": 3.5118, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 5.267815556590591, | |
| "grad_norm": 0.32533538341522217, | |
| "learning_rate": 0.0005371030867792661, | |
| "loss": 3.5024, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 5.282370749883558, | |
| "grad_norm": 0.3238582909107208, | |
| "learning_rate": 0.0005369283634245778, | |
| "loss": 3.5013, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 5.296925943176525, | |
| "grad_norm": 0.3118095397949219, | |
| "learning_rate": 0.0005367536400698893, | |
| "loss": 3.5033, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 5.311481136469492, | |
| "grad_norm": 0.3017449676990509, | |
| "learning_rate": 0.0005365789167152008, | |
| "loss": 3.5173, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 5.326036329762459, | |
| "grad_norm": 0.330851674079895, | |
| "learning_rate": 0.0005364041933605125, | |
| "loss": 3.5212, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 5.340591523055426, | |
| "grad_norm": 0.3320919871330261, | |
| "learning_rate": 0.000536229470005824, | |
| "loss": 3.5108, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 5.3551467163483935, | |
| "grad_norm": 0.3322393000125885, | |
| "learning_rate": 0.0005360547466511357, | |
| "loss": 3.5098, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 5.36970190964136, | |
| "grad_norm": 0.309768408536911, | |
| "learning_rate": 0.0005358800232964472, | |
| "loss": 3.5267, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 5.384257102934327, | |
| "grad_norm": 0.3261014223098755, | |
| "learning_rate": 0.0005357052999417589, | |
| "loss": 3.5207, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 5.398812296227294, | |
| "grad_norm": 0.322579562664032, | |
| "learning_rate": 0.0005355305765870704, | |
| "loss": 3.5036, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 5.413367489520261, | |
| "grad_norm": 0.32647162675857544, | |
| "learning_rate": 0.000535355853232382, | |
| "loss": 3.5241, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 5.427922682813228, | |
| "grad_norm": 0.31812402606010437, | |
| "learning_rate": 0.0005351811298776936, | |
| "loss": 3.5198, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 5.442477876106195, | |
| "grad_norm": 0.3058462142944336, | |
| "learning_rate": 0.0005350064065230052, | |
| "loss": 3.5234, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 5.457033069399162, | |
| "grad_norm": 0.3365894854068756, | |
| "learning_rate": 0.0005348316831683168, | |
| "loss": 3.5197, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 5.471588262692128, | |
| "grad_norm": 0.32743752002716064, | |
| "learning_rate": 0.0005346569598136284, | |
| "loss": 3.5149, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 5.486143455985095, | |
| "grad_norm": 0.32605621218681335, | |
| "learning_rate": 0.00053448223645894, | |
| "loss": 3.5251, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 5.500698649278062, | |
| "grad_norm": 0.3102964758872986, | |
| "learning_rate": 0.0005343075131042515, | |
| "loss": 3.526, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 5.515253842571029, | |
| "grad_norm": 0.2993220090866089, | |
| "learning_rate": 0.0005341327897495632, | |
| "loss": 3.5234, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 5.529809035863996, | |
| "grad_norm": 0.31743472814559937, | |
| "learning_rate": 0.0005339580663948748, | |
| "loss": 3.5155, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.529809035863996, | |
| "eval_accuracy": 0.3610857219484946, | |
| "eval_loss": 3.626174211502075, | |
| "eval_runtime": 82.3065, | |
| "eval_samples_per_second": 202.305, | |
| "eval_steps_per_second": 12.648, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.5443642291569635, | |
| "grad_norm": 0.32072290778160095, | |
| "learning_rate": 0.0005337833430401863, | |
| "loss": 3.5203, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 5.5589194224499305, | |
| "grad_norm": 0.32108786702156067, | |
| "learning_rate": 0.0005336086196854979, | |
| "loss": 3.524, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 5.573474615742897, | |
| "grad_norm": 0.31697359681129456, | |
| "learning_rate": 0.0005334338963308095, | |
| "loss": 3.5182, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 5.588029809035864, | |
| "grad_norm": 0.34002214670181274, | |
| "learning_rate": 0.0005332591729761211, | |
| "loss": 3.5445, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 5.602585002328831, | |
| "grad_norm": 0.3232363164424896, | |
| "learning_rate": 0.0005330844496214327, | |
| "loss": 3.5179, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 5.617140195621798, | |
| "grad_norm": 0.33068376779556274, | |
| "learning_rate": 0.0005329097262667443, | |
| "loss": 3.5267, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 5.631695388914765, | |
| "grad_norm": 0.2958562970161438, | |
| "learning_rate": 0.0005327350029120559, | |
| "loss": 3.5314, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 5.646250582207732, | |
| "grad_norm": 0.30992650985717773, | |
| "learning_rate": 0.0005325602795573674, | |
| "loss": 3.5335, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 5.660805775500698, | |
| "grad_norm": 0.32832181453704834, | |
| "learning_rate": 0.000532385556202679, | |
| "loss": 3.5311, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 5.675360968793665, | |
| "grad_norm": 0.29766660928726196, | |
| "learning_rate": 0.0005322108328479906, | |
| "loss": 3.52, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 5.689916162086632, | |
| "grad_norm": 0.3477511405944824, | |
| "learning_rate": 0.0005320361094933023, | |
| "loss": 3.5272, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 5.704471355379599, | |
| "grad_norm": 0.3183441460132599, | |
| "learning_rate": 0.0005318613861386138, | |
| "loss": 3.5206, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 5.719026548672566, | |
| "grad_norm": 0.3065016269683838, | |
| "learning_rate": 0.0005316866627839254, | |
| "loss": 3.5232, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 5.7335817419655335, | |
| "grad_norm": 0.30901676416397095, | |
| "learning_rate": 0.000531511939429237, | |
| "loss": 3.5229, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 5.748136935258501, | |
| "grad_norm": 0.3254716992378235, | |
| "learning_rate": 0.0005313372160745486, | |
| "loss": 3.5287, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 5.762692128551468, | |
| "grad_norm": 0.3171404004096985, | |
| "learning_rate": 0.0005311624927198602, | |
| "loss": 3.5372, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 5.777247321844434, | |
| "grad_norm": 0.31403276324272156, | |
| "learning_rate": 0.0005309877693651717, | |
| "loss": 3.5231, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 5.791802515137401, | |
| "grad_norm": 0.33173394203186035, | |
| "learning_rate": 0.0005308130460104834, | |
| "loss": 3.5251, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 5.806357708430368, | |
| "grad_norm": 0.309815376996994, | |
| "learning_rate": 0.0005306383226557949, | |
| "loss": 3.5227, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 5.820912901723335, | |
| "grad_norm": 0.31985828280448914, | |
| "learning_rate": 0.0005304635993011065, | |
| "loss": 3.5286, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.820912901723335, | |
| "eval_accuracy": 0.3622141016077618, | |
| "eval_loss": 3.6125025749206543, | |
| "eval_runtime": 82.4287, | |
| "eval_samples_per_second": 202.005, | |
| "eval_steps_per_second": 12.629, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.835468095016302, | |
| "grad_norm": 0.3527766466140747, | |
| "learning_rate": 0.0005302888759464181, | |
| "loss": 3.5315, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 5.850023288309269, | |
| "grad_norm": 0.33389145135879517, | |
| "learning_rate": 0.0005301141525917298, | |
| "loss": 3.5286, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 5.864578481602235, | |
| "grad_norm": 0.3075472116470337, | |
| "learning_rate": 0.0005299394292370413, | |
| "loss": 3.5311, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 5.879133674895202, | |
| "grad_norm": 0.31657978892326355, | |
| "learning_rate": 0.0005297647058823528, | |
| "loss": 3.5477, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 5.893688868188169, | |
| "grad_norm": 0.3266170620918274, | |
| "learning_rate": 0.0005295899825276645, | |
| "loss": 3.5205, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 5.9082440614811365, | |
| "grad_norm": 0.3099294900894165, | |
| "learning_rate": 0.000529415259172976, | |
| "loss": 3.528, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 5.9227992547741035, | |
| "grad_norm": 0.33021819591522217, | |
| "learning_rate": 0.0005292405358182877, | |
| "loss": 3.5317, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 5.937354448067071, | |
| "grad_norm": 0.34638336300849915, | |
| "learning_rate": 0.0005290658124635992, | |
| "loss": 3.5278, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 5.951909641360038, | |
| "grad_norm": 0.3213142156600952, | |
| "learning_rate": 0.0005288910891089109, | |
| "loss": 3.5322, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 5.966464834653004, | |
| "grad_norm": 0.3418400287628174, | |
| "learning_rate": 0.0005287163657542224, | |
| "loss": 3.5361, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 5.981020027945971, | |
| "grad_norm": 0.3340821862220764, | |
| "learning_rate": 0.000528541642399534, | |
| "loss": 3.5307, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 5.995575221238938, | |
| "grad_norm": 0.3119957149028778, | |
| "learning_rate": 0.0005283669190448456, | |
| "loss": 3.5314, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 6.009897531439218, | |
| "grad_norm": 0.32535383105278015, | |
| "learning_rate": 0.0005281921956901572, | |
| "loss": 3.4512, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 6.024452724732185, | |
| "grad_norm": 0.3190499544143677, | |
| "learning_rate": 0.0005280174723354688, | |
| "loss": 3.4188, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 6.039007918025152, | |
| "grad_norm": 0.33317360281944275, | |
| "learning_rate": 0.0005278427489807804, | |
| "loss": 3.433, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 6.053563111318118, | |
| "grad_norm": 0.3085060119628906, | |
| "learning_rate": 0.000527668025626092, | |
| "loss": 3.4306, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 6.068118304611085, | |
| "grad_norm": 0.33972272276878357, | |
| "learning_rate": 0.0005274933022714035, | |
| "loss": 3.4291, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 6.082673497904052, | |
| "grad_norm": 0.3487497568130493, | |
| "learning_rate": 0.0005273185789167152, | |
| "loss": 3.4304, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 6.097228691197019, | |
| "grad_norm": 0.3195396959781647, | |
| "learning_rate": 0.0005271438555620268, | |
| "loss": 3.4416, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 6.111783884489986, | |
| "grad_norm": 0.32456278800964355, | |
| "learning_rate": 0.0005269691322073384, | |
| "loss": 3.4429, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.111783884489986, | |
| "eval_accuracy": 0.36271488545612524, | |
| "eval_loss": 3.6173150539398193, | |
| "eval_runtime": 82.2876, | |
| "eval_samples_per_second": 202.351, | |
| "eval_steps_per_second": 12.651, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.126339077782953, | |
| "grad_norm": 0.31698915362358093, | |
| "learning_rate": 0.0005267944088526499, | |
| "loss": 3.4385, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 6.14089427107592, | |
| "grad_norm": 0.3022404611110687, | |
| "learning_rate": 0.0005266196854979615, | |
| "loss": 3.4463, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 6.155449464368886, | |
| "grad_norm": 0.316621869802475, | |
| "learning_rate": 0.0005264449621432731, | |
| "loss": 3.4348, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 6.1700046576618535, | |
| "grad_norm": 0.3234342038631439, | |
| "learning_rate": 0.0005262702387885847, | |
| "loss": 3.4546, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 6.1845598509548205, | |
| "grad_norm": 0.322392076253891, | |
| "learning_rate": 0.0005260955154338963, | |
| "loss": 3.4599, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 6.199115044247788, | |
| "grad_norm": 0.3075820207595825, | |
| "learning_rate": 0.0005259207920792079, | |
| "loss": 3.4521, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 6.213670237540755, | |
| "grad_norm": 0.3221750557422638, | |
| "learning_rate": 0.0005257460687245195, | |
| "loss": 3.4675, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 6.228225430833722, | |
| "grad_norm": 0.32317793369293213, | |
| "learning_rate": 0.000525571345369831, | |
| "loss": 3.4651, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 6.242780624126689, | |
| "grad_norm": 0.30188095569610596, | |
| "learning_rate": 0.0005253966220151426, | |
| "loss": 3.4652, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 6.257335817419655, | |
| "grad_norm": 0.3238082230091095, | |
| "learning_rate": 0.0005252218986604543, | |
| "loss": 3.4693, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 6.271891010712622, | |
| "grad_norm": 0.313158243894577, | |
| "learning_rate": 0.0005250471753057658, | |
| "loss": 3.4643, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 6.286446204005589, | |
| "grad_norm": 0.3327135145664215, | |
| "learning_rate": 0.0005248724519510774, | |
| "loss": 3.4673, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 6.301001397298556, | |
| "grad_norm": 0.3160534203052521, | |
| "learning_rate": 0.000524697728596389, | |
| "loss": 3.4619, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 6.315556590591523, | |
| "grad_norm": 0.3262614607810974, | |
| "learning_rate": 0.0005245230052417006, | |
| "loss": 3.4656, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 6.33011178388449, | |
| "grad_norm": 0.3326106071472168, | |
| "learning_rate": 0.0005243482818870122, | |
| "loss": 3.472, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 6.344666977177457, | |
| "grad_norm": 0.3145747482776642, | |
| "learning_rate": 0.0005241735585323238, | |
| "loss": 3.4845, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 6.3592221704704235, | |
| "grad_norm": 0.34462374448776245, | |
| "learning_rate": 0.0005239988351776354, | |
| "loss": 3.4617, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 6.3737773637633905, | |
| "grad_norm": 0.3171409070491791, | |
| "learning_rate": 0.0005238241118229469, | |
| "loss": 3.4771, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 6.388332557056358, | |
| "grad_norm": 0.332462877035141, | |
| "learning_rate": 0.0005236493884682585, | |
| "loss": 3.4645, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 6.402887750349325, | |
| "grad_norm": 0.32564809918403625, | |
| "learning_rate": 0.0005234746651135701, | |
| "loss": 3.4666, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.402887750349325, | |
| "eval_accuracy": 0.3636495801160723, | |
| "eval_loss": 3.6089110374450684, | |
| "eval_runtime": 82.3185, | |
| "eval_samples_per_second": 202.275, | |
| "eval_steps_per_second": 12.646, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.417442943642292, | |
| "grad_norm": 0.32464858889579773, | |
| "learning_rate": 0.0005232999417588818, | |
| "loss": 3.4765, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 6.431998136935259, | |
| "grad_norm": 0.30988913774490356, | |
| "learning_rate": 0.0005231252184041933, | |
| "loss": 3.4713, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 6.446553330228225, | |
| "grad_norm": 0.31231069564819336, | |
| "learning_rate": 0.0005229504950495049, | |
| "loss": 3.4653, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 6.461108523521192, | |
| "grad_norm": 0.3273122012615204, | |
| "learning_rate": 0.0005227757716948165, | |
| "loss": 3.4655, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 6.475663716814159, | |
| "grad_norm": 0.3160896301269531, | |
| "learning_rate": 0.000522601048340128, | |
| "loss": 3.4772, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 6.490218910107126, | |
| "grad_norm": 0.31845608353614807, | |
| "learning_rate": 0.0005224263249854397, | |
| "loss": 3.4714, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 6.504774103400093, | |
| "grad_norm": 0.3098876178264618, | |
| "learning_rate": 0.0005222516016307512, | |
| "loss": 3.4904, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 6.51932929669306, | |
| "grad_norm": 0.302150160074234, | |
| "learning_rate": 0.0005220768782760629, | |
| "loss": 3.4713, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 6.533884489986027, | |
| "grad_norm": 0.3412488102912903, | |
| "learning_rate": 0.0005219021549213744, | |
| "loss": 3.48, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 6.548439683278994, | |
| "grad_norm": 0.31061360239982605, | |
| "learning_rate": 0.000521727431566686, | |
| "loss": 3.4889, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 6.562994876571961, | |
| "grad_norm": 0.3276578187942505, | |
| "learning_rate": 0.0005215527082119976, | |
| "loss": 3.4738, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 6.577550069864928, | |
| "grad_norm": 0.3279768228530884, | |
| "learning_rate": 0.0005213779848573093, | |
| "loss": 3.4726, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 6.592105263157895, | |
| "grad_norm": 0.33064624667167664, | |
| "learning_rate": 0.0005212032615026208, | |
| "loss": 3.4793, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 6.606660456450862, | |
| "grad_norm": 0.3199060559272766, | |
| "learning_rate": 0.0005210285381479323, | |
| "loss": 3.4794, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 6.621215649743829, | |
| "grad_norm": 0.31415632367134094, | |
| "learning_rate": 0.000520853814793244, | |
| "loss": 3.5016, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 6.635770843036796, | |
| "grad_norm": 0.324936181306839, | |
| "learning_rate": 0.0005206790914385555, | |
| "loss": 3.4926, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 6.650326036329762, | |
| "grad_norm": 0.34248557686805725, | |
| "learning_rate": 0.0005205043680838672, | |
| "loss": 3.4842, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 6.664881229622729, | |
| "grad_norm": 0.3163456618785858, | |
| "learning_rate": 0.0005203296447291787, | |
| "loss": 3.4787, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 6.679436422915696, | |
| "grad_norm": 0.36227306723594666, | |
| "learning_rate": 0.0005201549213744904, | |
| "loss": 3.5002, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 6.693991616208663, | |
| "grad_norm": 0.3207639455795288, | |
| "learning_rate": 0.0005199801980198019, | |
| "loss": 3.4875, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.693991616208663, | |
| "eval_accuracy": 0.36408725180142915, | |
| "eval_loss": 3.5998013019561768, | |
| "eval_runtime": 82.2909, | |
| "eval_samples_per_second": 202.343, | |
| "eval_steps_per_second": 12.65, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.70854680950163, | |
| "grad_norm": 0.32826781272888184, | |
| "learning_rate": 0.0005198054746651136, | |
| "loss": 3.5005, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 6.723102002794597, | |
| "grad_norm": 0.32250669598579407, | |
| "learning_rate": 0.0005196307513104251, | |
| "loss": 3.4902, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 6.737657196087564, | |
| "grad_norm": 0.3102186918258667, | |
| "learning_rate": 0.0005194560279557367, | |
| "loss": 3.4858, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 6.752212389380531, | |
| "grad_norm": 0.3280225396156311, | |
| "learning_rate": 0.0005192813046010483, | |
| "loss": 3.4815, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 6.766767582673498, | |
| "grad_norm": 0.3475222587585449, | |
| "learning_rate": 0.0005191065812463599, | |
| "loss": 3.4891, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 6.781322775966465, | |
| "grad_norm": 0.32275667786598206, | |
| "learning_rate": 0.0005189318578916715, | |
| "loss": 3.4849, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 6.795877969259432, | |
| "grad_norm": 0.33388715982437134, | |
| "learning_rate": 0.000518757134536983, | |
| "loss": 3.4788, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 6.810433162552399, | |
| "grad_norm": 0.33769869804382324, | |
| "learning_rate": 0.0005185824111822947, | |
| "loss": 3.4834, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 6.824988355845366, | |
| "grad_norm": 0.3156081736087799, | |
| "learning_rate": 0.0005184076878276063, | |
| "loss": 3.4717, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 6.839543549138332, | |
| "grad_norm": 0.310829222202301, | |
| "learning_rate": 0.0005182329644729179, | |
| "loss": 3.5004, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 6.854098742431299, | |
| "grad_norm": 0.321531742811203, | |
| "learning_rate": 0.0005180582411182294, | |
| "loss": 3.4872, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 6.868653935724266, | |
| "grad_norm": 0.33255383372306824, | |
| "learning_rate": 0.000517883517763541, | |
| "loss": 3.5063, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 6.883209129017233, | |
| "grad_norm": 0.31028273701667786, | |
| "learning_rate": 0.0005177087944088526, | |
| "loss": 3.5077, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 6.8977643223102, | |
| "grad_norm": 0.31035181879997253, | |
| "learning_rate": 0.0005175340710541642, | |
| "loss": 3.4866, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 6.912319515603167, | |
| "grad_norm": 0.31630781292915344, | |
| "learning_rate": 0.0005173593476994758, | |
| "loss": 3.4924, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 6.926874708896134, | |
| "grad_norm": 0.31600630283355713, | |
| "learning_rate": 0.0005171846243447874, | |
| "loss": 3.4943, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 6.9414299021891015, | |
| "grad_norm": 0.326752245426178, | |
| "learning_rate": 0.000517009900990099, | |
| "loss": 3.4939, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 6.955985095482068, | |
| "grad_norm": 0.295224666595459, | |
| "learning_rate": 0.0005168351776354105, | |
| "loss": 3.484, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 6.970540288775035, | |
| "grad_norm": 0.3208736479282379, | |
| "learning_rate": 0.0005166604542807221, | |
| "loss": 3.5007, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 6.985095482068002, | |
| "grad_norm": 0.317485511302948, | |
| "learning_rate": 0.0005164857309260338, | |
| "loss": 3.4958, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.985095482068002, | |
| "eval_accuracy": 0.36497646339418155, | |
| "eval_loss": 3.5896191596984863, | |
| "eval_runtime": 82.3792, | |
| "eval_samples_per_second": 202.126, | |
| "eval_steps_per_second": 12.637, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.999650675360969, | |
| "grad_norm": 0.3053331673145294, | |
| "learning_rate": 0.0005163110075713453, | |
| "loss": 3.4978, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 7.0139729855612485, | |
| "grad_norm": 0.3304654061794281, | |
| "learning_rate": 0.0005161362842166569, | |
| "loss": 3.3864, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 7.0285281788542155, | |
| "grad_norm": 0.3205164074897766, | |
| "learning_rate": 0.0005159615608619685, | |
| "loss": 3.3796, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 7.043083372147182, | |
| "grad_norm": 0.34038031101226807, | |
| "learning_rate": 0.0005157868375072801, | |
| "loss": 3.384, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 7.057638565440149, | |
| "grad_norm": 0.331194669008255, | |
| "learning_rate": 0.0005156121141525917, | |
| "loss": 3.3836, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 7.072193758733116, | |
| "grad_norm": 0.32286176085472107, | |
| "learning_rate": 0.0005154373907979033, | |
| "loss": 3.3911, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 7.086748952026083, | |
| "grad_norm": 0.32159680128097534, | |
| "learning_rate": 0.0005152626674432149, | |
| "loss": 3.3887, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 7.10130414531905, | |
| "grad_norm": 0.34153318405151367, | |
| "learning_rate": 0.0005150879440885264, | |
| "loss": 3.4101, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 7.115859338612017, | |
| "grad_norm": 0.3122780919075012, | |
| "learning_rate": 0.000514913220733838, | |
| "loss": 3.4079, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 7.130414531904984, | |
| "grad_norm": 0.3258710503578186, | |
| "learning_rate": 0.0005147384973791496, | |
| "loss": 3.3973, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 7.14496972519795, | |
| "grad_norm": 0.33584776520729065, | |
| "learning_rate": 0.0005145637740244613, | |
| "loss": 3.4094, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 7.159524918490917, | |
| "grad_norm": 0.31355953216552734, | |
| "learning_rate": 0.0005143890506697728, | |
| "loss": 3.4092, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 7.174080111783884, | |
| "grad_norm": 0.32509100437164307, | |
| "learning_rate": 0.0005142143273150844, | |
| "loss": 3.4189, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 7.1886353050768514, | |
| "grad_norm": 0.3257924020290375, | |
| "learning_rate": 0.000514039603960396, | |
| "loss": 3.4149, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 7.2031904983698185, | |
| "grad_norm": 0.32637664675712585, | |
| "learning_rate": 0.0005138648806057075, | |
| "loss": 3.4279, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 7.217745691662786, | |
| "grad_norm": 0.33568671345710754, | |
| "learning_rate": 0.0005136901572510192, | |
| "loss": 3.4036, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 7.232300884955753, | |
| "grad_norm": 0.33469074964523315, | |
| "learning_rate": 0.0005135154338963307, | |
| "loss": 3.4213, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 7.246856078248719, | |
| "grad_norm": 0.33833402395248413, | |
| "learning_rate": 0.0005133407105416424, | |
| "loss": 3.4239, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 7.261411271541686, | |
| "grad_norm": 0.33621230721473694, | |
| "learning_rate": 0.0005131659871869539, | |
| "loss": 3.4226, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 7.275966464834653, | |
| "grad_norm": 0.32759207487106323, | |
| "learning_rate": 0.0005129912638322656, | |
| "loss": 3.4263, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.275966464834653, | |
| "eval_accuracy": 0.3646007285987772, | |
| "eval_loss": 3.5983967781066895, | |
| "eval_runtime": 82.293, | |
| "eval_samples_per_second": 202.338, | |
| "eval_steps_per_second": 12.65, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.29052165812762, | |
| "grad_norm": 0.3359229862689972, | |
| "learning_rate": 0.0005128165404775771, | |
| "loss": 3.4338, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 7.305076851420587, | |
| "grad_norm": 0.32089221477508545, | |
| "learning_rate": 0.0005126418171228888, | |
| "loss": 3.4223, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 7.319632044713554, | |
| "grad_norm": 0.3113846480846405, | |
| "learning_rate": 0.0005124670937682003, | |
| "loss": 3.4302, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 7.334187238006521, | |
| "grad_norm": 0.3196338415145874, | |
| "learning_rate": 0.000512292370413512, | |
| "loss": 3.4346, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 7.348742431299487, | |
| "grad_norm": 0.33884793519973755, | |
| "learning_rate": 0.0005121176470588235, | |
| "loss": 3.4392, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 7.363297624592454, | |
| "grad_norm": 0.33437052369117737, | |
| "learning_rate": 0.000511942923704135, | |
| "loss": 3.4403, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 7.3778528178854215, | |
| "grad_norm": 0.33692166209220886, | |
| "learning_rate": 0.0005117682003494467, | |
| "loss": 3.4454, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 7.3924080111783885, | |
| "grad_norm": 0.34315139055252075, | |
| "learning_rate": 0.0005115934769947583, | |
| "loss": 3.4474, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 7.406963204471356, | |
| "grad_norm": 0.3370000720024109, | |
| "learning_rate": 0.0005114187536400699, | |
| "loss": 3.439, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 7.421518397764323, | |
| "grad_norm": 0.31308624148368835, | |
| "learning_rate": 0.0005112440302853814, | |
| "loss": 3.4302, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 7.436073591057289, | |
| "grad_norm": 0.3229582607746124, | |
| "learning_rate": 0.0005110693069306931, | |
| "loss": 3.4391, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 7.450628784350256, | |
| "grad_norm": 0.5394797325134277, | |
| "learning_rate": 0.0005108945835760046, | |
| "loss": 3.4402, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 7.465183977643223, | |
| "grad_norm": 0.32158133387565613, | |
| "learning_rate": 0.0005107198602213162, | |
| "loss": 3.4453, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 7.47973917093619, | |
| "grad_norm": 0.33943256735801697, | |
| "learning_rate": 0.0005105451368666278, | |
| "loss": 3.4401, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 7.494294364229157, | |
| "grad_norm": 0.3760105073451996, | |
| "learning_rate": 0.0005103704135119394, | |
| "loss": 3.4552, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 7.508849557522124, | |
| "grad_norm": 0.3264046907424927, | |
| "learning_rate": 0.000510195690157251, | |
| "loss": 3.4343, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 7.523404750815091, | |
| "grad_norm": 0.3396759331226349, | |
| "learning_rate": 0.0005100209668025625, | |
| "loss": 3.4502, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 7.537959944108058, | |
| "grad_norm": 0.34846237301826477, | |
| "learning_rate": 0.0005098462434478742, | |
| "loss": 3.4549, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 7.552515137401024, | |
| "grad_norm": 0.3175800144672394, | |
| "learning_rate": 0.0005096715200931858, | |
| "loss": 3.4398, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 7.5670703306939915, | |
| "grad_norm": 0.31833750009536743, | |
| "learning_rate": 0.0005094967967384974, | |
| "loss": 3.445, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.5670703306939915, | |
| "eval_accuracy": 0.36580068238704067, | |
| "eval_loss": 3.590224027633667, | |
| "eval_runtime": 82.2758, | |
| "eval_samples_per_second": 202.38, | |
| "eval_steps_per_second": 12.653, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.5816255239869585, | |
| "grad_norm": 0.3222794532775879, | |
| "learning_rate": 0.0005093220733838089, | |
| "loss": 3.45, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 7.596180717279926, | |
| "grad_norm": 0.3188936710357666, | |
| "learning_rate": 0.0005091473500291205, | |
| "loss": 3.4412, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 7.610735910572893, | |
| "grad_norm": 0.3307031989097595, | |
| "learning_rate": 0.0005089726266744321, | |
| "loss": 3.4559, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 7.625291103865859, | |
| "grad_norm": 0.31371206045150757, | |
| "learning_rate": 0.0005087979033197437, | |
| "loss": 3.4461, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 7.639846297158826, | |
| "grad_norm": 0.3318116068840027, | |
| "learning_rate": 0.0005086231799650553, | |
| "loss": 3.4565, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 7.654401490451793, | |
| "grad_norm": 0.32326042652130127, | |
| "learning_rate": 0.0005084484566103669, | |
| "loss": 3.4577, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 7.66895668374476, | |
| "grad_norm": 0.33533239364624023, | |
| "learning_rate": 0.0005082737332556785, | |
| "loss": 3.455, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 7.683511877037727, | |
| "grad_norm": 0.3287736177444458, | |
| "learning_rate": 0.00050809900990099, | |
| "loss": 3.4567, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 7.698067070330694, | |
| "grad_norm": 0.31383323669433594, | |
| "learning_rate": 0.0005079242865463016, | |
| "loss": 3.4535, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 7.712622263623661, | |
| "grad_norm": 0.33366987109184265, | |
| "learning_rate": 0.0005077495631916133, | |
| "loss": 3.4511, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 7.727177456916628, | |
| "grad_norm": 0.3248838484287262, | |
| "learning_rate": 0.0005075748398369248, | |
| "loss": 3.4487, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 7.7417326502095944, | |
| "grad_norm": 0.3126499652862549, | |
| "learning_rate": 0.0005074001164822364, | |
| "loss": 3.4647, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 7.7562878435025615, | |
| "grad_norm": 0.35184940695762634, | |
| "learning_rate": 0.000507225393127548, | |
| "loss": 3.4539, | |
| "step": 26650 | |
| }, | |
| { | |
| "epoch": 7.770843036795529, | |
| "grad_norm": 0.3201959729194641, | |
| "learning_rate": 0.0005070506697728596, | |
| "loss": 3.4597, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 7.785398230088496, | |
| "grad_norm": 0.33270755410194397, | |
| "learning_rate": 0.0005068759464181711, | |
| "loss": 3.4628, | |
| "step": 26750 | |
| }, | |
| { | |
| "epoch": 7.799953423381463, | |
| "grad_norm": 0.3396896421909332, | |
| "learning_rate": 0.0005067012230634828, | |
| "loss": 3.4634, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 7.81450861667443, | |
| "grad_norm": 0.32018783688545227, | |
| "learning_rate": 0.0005065264997087944, | |
| "loss": 3.4568, | |
| "step": 26850 | |
| }, | |
| { | |
| "epoch": 7.829063809967396, | |
| "grad_norm": 0.31735870242118835, | |
| "learning_rate": 0.0005063517763541059, | |
| "loss": 3.4532, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 7.843619003260363, | |
| "grad_norm": 0.3333510458469391, | |
| "learning_rate": 0.0005061770529994175, | |
| "loss": 3.4689, | |
| "step": 26950 | |
| }, | |
| { | |
| "epoch": 7.85817419655333, | |
| "grad_norm": 0.3392159044742584, | |
| "learning_rate": 0.0005060023296447291, | |
| "loss": 3.4703, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.85817419655333, | |
| "eval_accuracy": 0.3662585687689285, | |
| "eval_loss": 3.5806941986083984, | |
| "eval_runtime": 82.3938, | |
| "eval_samples_per_second": 202.091, | |
| "eval_steps_per_second": 12.634, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.872729389846297, | |
| "grad_norm": 0.3154710531234741, | |
| "learning_rate": 0.0005058276062900408, | |
| "loss": 3.4538, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 7.887284583139264, | |
| "grad_norm": 0.3323320746421814, | |
| "learning_rate": 0.0005056528829353523, | |
| "loss": 3.458, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 7.901839776432231, | |
| "grad_norm": 0.30794772505760193, | |
| "learning_rate": 0.000505478159580664, | |
| "loss": 3.4727, | |
| "step": 27150 | |
| }, | |
| { | |
| "epoch": 7.916394969725198, | |
| "grad_norm": 0.3046403229236603, | |
| "learning_rate": 0.0005053034362259755, | |
| "loss": 3.4543, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 7.930950163018165, | |
| "grad_norm": 0.3160116970539093, | |
| "learning_rate": 0.000505128712871287, | |
| "loss": 3.45, | |
| "step": 27250 | |
| }, | |
| { | |
| "epoch": 7.9455053563111315, | |
| "grad_norm": 0.3044631779193878, | |
| "learning_rate": 0.0005049539895165987, | |
| "loss": 3.4675, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 7.960060549604099, | |
| "grad_norm": 0.3197316825389862, | |
| "learning_rate": 0.0005047792661619103, | |
| "loss": 3.4633, | |
| "step": 27350 | |
| }, | |
| { | |
| "epoch": 7.974615742897066, | |
| "grad_norm": 0.32045426964759827, | |
| "learning_rate": 0.0005046045428072219, | |
| "loss": 3.4549, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 7.989170936190033, | |
| "grad_norm": 0.32003840804100037, | |
| "learning_rate": 0.0005044298194525334, | |
| "loss": 3.4735, | |
| "step": 27450 | |
| }, | |
| { | |
| "epoch": 8.003493246390311, | |
| "grad_norm": 0.329167902469635, | |
| "learning_rate": 0.0005042550960978451, | |
| "loss": 3.4404, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 8.018048439683279, | |
| "grad_norm": 0.3276357352733612, | |
| "learning_rate": 0.0005040803727431566, | |
| "loss": 3.3534, | |
| "step": 27550 | |
| }, | |
| { | |
| "epoch": 8.032603632976246, | |
| "grad_norm": 0.33727550506591797, | |
| "learning_rate": 0.0005039056493884683, | |
| "loss": 3.3633, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 8.047158826269213, | |
| "grad_norm": 0.321049302816391, | |
| "learning_rate": 0.0005037309260337798, | |
| "loss": 3.3465, | |
| "step": 27650 | |
| }, | |
| { | |
| "epoch": 8.06171401956218, | |
| "grad_norm": 0.3481226861476898, | |
| "learning_rate": 0.0005035562026790914, | |
| "loss": 3.3578, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 8.076269212855147, | |
| "grad_norm": 0.3485446274280548, | |
| "learning_rate": 0.000503381479324403, | |
| "loss": 3.378, | |
| "step": 27750 | |
| }, | |
| { | |
| "epoch": 8.090824406148114, | |
| "grad_norm": 0.32518789172172546, | |
| "learning_rate": 0.0005032067559697145, | |
| "loss": 3.3601, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 8.10537959944108, | |
| "grad_norm": 0.3377379775047302, | |
| "learning_rate": 0.0005030320326150262, | |
| "loss": 3.3731, | |
| "step": 27850 | |
| }, | |
| { | |
| "epoch": 8.119934792734048, | |
| "grad_norm": 0.33603596687316895, | |
| "learning_rate": 0.0005028573092603378, | |
| "loss": 3.372, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 8.134489986027015, | |
| "grad_norm": 0.355691134929657, | |
| "learning_rate": 0.0005026825859056494, | |
| "loss": 3.379, | |
| "step": 27950 | |
| }, | |
| { | |
| "epoch": 8.149045179319982, | |
| "grad_norm": 0.3463926613330841, | |
| "learning_rate": 0.0005025078625509609, | |
| "loss": 3.3754, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.149045179319982, | |
| "eval_accuracy": 0.3661492683748947, | |
| "eval_loss": 3.5915322303771973, | |
| "eval_runtime": 82.2845, | |
| "eval_samples_per_second": 202.359, | |
| "eval_steps_per_second": 12.651, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.16360037261295, | |
| "grad_norm": 0.32895907759666443, | |
| "learning_rate": 0.0005023331391962726, | |
| "loss": 3.3723, | |
| "step": 28050 | |
| }, | |
| { | |
| "epoch": 8.178155565905914, | |
| "grad_norm": 0.3400669991970062, | |
| "learning_rate": 0.0005021584158415841, | |
| "loss": 3.3726, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 8.192710759198881, | |
| "grad_norm": 0.36708810925483704, | |
| "learning_rate": 0.0005019836924868956, | |
| "loss": 3.3925, | |
| "step": 28150 | |
| }, | |
| { | |
| "epoch": 8.207265952491849, | |
| "grad_norm": 0.3426344096660614, | |
| "learning_rate": 0.0005018089691322073, | |
| "loss": 3.3747, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 8.221821145784816, | |
| "grad_norm": 0.33265358209609985, | |
| "learning_rate": 0.0005016342457775189, | |
| "loss": 3.3824, | |
| "step": 28250 | |
| }, | |
| { | |
| "epoch": 8.236376339077783, | |
| "grad_norm": 0.33101364970207214, | |
| "learning_rate": 0.0005014595224228305, | |
| "loss": 3.3913, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 8.25093153237075, | |
| "grad_norm": 0.32427892088890076, | |
| "learning_rate": 0.000501284799068142, | |
| "loss": 3.4026, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 8.265486725663717, | |
| "grad_norm": 0.32167622447013855, | |
| "learning_rate": 0.0005011100757134537, | |
| "loss": 3.3815, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 8.280041918956684, | |
| "grad_norm": 0.3368193507194519, | |
| "learning_rate": 0.0005009353523587653, | |
| "loss": 3.3923, | |
| "step": 28450 | |
| }, | |
| { | |
| "epoch": 8.294597112249651, | |
| "grad_norm": 0.3301495313644409, | |
| "learning_rate": 0.0005007606290040768, | |
| "loss": 3.4069, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 8.309152305542618, | |
| "grad_norm": 0.3171535134315491, | |
| "learning_rate": 0.0005005859056493884, | |
| "loss": 3.4032, | |
| "step": 28550 | |
| }, | |
| { | |
| "epoch": 8.323707498835585, | |
| "grad_norm": 0.3733464181423187, | |
| "learning_rate": 0.0005004111822947, | |
| "loss": 3.4099, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 8.338262692128552, | |
| "grad_norm": 0.3598664402961731, | |
| "learning_rate": 0.0005002364589400116, | |
| "loss": 3.4107, | |
| "step": 28650 | |
| }, | |
| { | |
| "epoch": 8.35281788542152, | |
| "grad_norm": 0.31014010310173035, | |
| "learning_rate": 0.0005000617355853231, | |
| "loss": 3.3965, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 8.367373078714486, | |
| "grad_norm": 0.3312365710735321, | |
| "learning_rate": 0.0004998870122306348, | |
| "loss": 3.4086, | |
| "step": 28750 | |
| }, | |
| { | |
| "epoch": 8.381928272007451, | |
| "grad_norm": 0.3404453992843628, | |
| "learning_rate": 0.0004997122888759464, | |
| "loss": 3.4187, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 8.396483465300419, | |
| "grad_norm": 0.32525068521499634, | |
| "learning_rate": 0.000499537565521258, | |
| "loss": 3.4138, | |
| "step": 28850 | |
| }, | |
| { | |
| "epoch": 8.411038658593386, | |
| "grad_norm": 0.3293299674987793, | |
| "learning_rate": 0.0004993628421665695, | |
| "loss": 3.4197, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 8.425593851886353, | |
| "grad_norm": 0.3383057713508606, | |
| "learning_rate": 0.0004991881188118811, | |
| "loss": 3.4084, | |
| "step": 28950 | |
| }, | |
| { | |
| "epoch": 8.44014904517932, | |
| "grad_norm": 0.3335309326648712, | |
| "learning_rate": 0.0004990133954571928, | |
| "loss": 3.4168, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.44014904517932, | |
| "eval_accuracy": 0.3668672426836608, | |
| "eval_loss": 3.585283041000366, | |
| "eval_runtime": 82.4708, | |
| "eval_samples_per_second": 201.902, | |
| "eval_steps_per_second": 12.623, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.454704238472287, | |
| "grad_norm": 0.3346988260746002, | |
| "learning_rate": 0.0004988386721025043, | |
| "loss": 3.42, | |
| "step": 29050 | |
| }, | |
| { | |
| "epoch": 8.469259431765254, | |
| "grad_norm": 0.31389665603637695, | |
| "learning_rate": 0.0004986639487478159, | |
| "loss": 3.4261, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 8.483814625058221, | |
| "grad_norm": 0.3408563435077667, | |
| "learning_rate": 0.0004984892253931275, | |
| "loss": 3.4063, | |
| "step": 29150 | |
| }, | |
| { | |
| "epoch": 8.498369818351188, | |
| "grad_norm": 0.3189634084701538, | |
| "learning_rate": 0.0004983145020384391, | |
| "loss": 3.4132, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 8.512925011644155, | |
| "grad_norm": 0.32136133313179016, | |
| "learning_rate": 0.0004981397786837507, | |
| "loss": 3.4198, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 8.527480204937122, | |
| "grad_norm": 0.32757824659347534, | |
| "learning_rate": 0.0004979650553290622, | |
| "loss": 3.4259, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 8.54203539823009, | |
| "grad_norm": 0.3273204565048218, | |
| "learning_rate": 0.0004977903319743739, | |
| "loss": 3.4243, | |
| "step": 29350 | |
| }, | |
| { | |
| "epoch": 8.556590591523056, | |
| "grad_norm": 0.31796106696128845, | |
| "learning_rate": 0.0004976156086196854, | |
| "loss": 3.4144, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 8.571145784816022, | |
| "grad_norm": 0.3298608064651489, | |
| "learning_rate": 0.0004974408852649971, | |
| "loss": 3.4266, | |
| "step": 29450 | |
| }, | |
| { | |
| "epoch": 8.585700978108989, | |
| "grad_norm": 0.3113996982574463, | |
| "learning_rate": 0.0004972661619103086, | |
| "loss": 3.4248, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 8.600256171401956, | |
| "grad_norm": 0.31293776631355286, | |
| "learning_rate": 0.0004970914385556202, | |
| "loss": 3.4089, | |
| "step": 29550 | |
| }, | |
| { | |
| "epoch": 8.614811364694923, | |
| "grad_norm": 0.32372698187828064, | |
| "learning_rate": 0.0004969167152009318, | |
| "loss": 3.4218, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 8.62936655798789, | |
| "grad_norm": 0.32352960109710693, | |
| "learning_rate": 0.0004967419918462435, | |
| "loss": 3.425, | |
| "step": 29650 | |
| }, | |
| { | |
| "epoch": 8.643921751280857, | |
| "grad_norm": 0.3247092068195343, | |
| "learning_rate": 0.000496567268491555, | |
| "loss": 3.4232, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 8.658476944573824, | |
| "grad_norm": 0.3242696523666382, | |
| "learning_rate": 0.0004963925451368665, | |
| "loss": 3.4211, | |
| "step": 29750 | |
| }, | |
| { | |
| "epoch": 8.673032137866791, | |
| "grad_norm": 0.3430579900741577, | |
| "learning_rate": 0.0004962178217821782, | |
| "loss": 3.4239, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 8.687587331159758, | |
| "grad_norm": 0.3495425879955292, | |
| "learning_rate": 0.0004960430984274898, | |
| "loss": 3.4195, | |
| "step": 29850 | |
| }, | |
| { | |
| "epoch": 8.702142524452725, | |
| "grad_norm": 0.3101871609687805, | |
| "learning_rate": 0.0004958683750728014, | |
| "loss": 3.4257, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 8.716697717745692, | |
| "grad_norm": 0.3443967401981354, | |
| "learning_rate": 0.0004956936517181129, | |
| "loss": 3.4379, | |
| "step": 29950 | |
| }, | |
| { | |
| "epoch": 8.73125291103866, | |
| "grad_norm": 0.3173809349536896, | |
| "learning_rate": 0.0004955189283634246, | |
| "loss": 3.4287, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.73125291103866, | |
| "eval_accuracy": 0.3671833911352209, | |
| "eval_loss": 3.574448823928833, | |
| "eval_runtime": 82.5274, | |
| "eval_samples_per_second": 201.763, | |
| "eval_steps_per_second": 12.614, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.745808104331626, | |
| "grad_norm": 0.33884406089782715, | |
| "learning_rate": 0.0004953442050087361, | |
| "loss": 3.4242, | |
| "step": 30050 | |
| }, | |
| { | |
| "epoch": 8.760363297624593, | |
| "grad_norm": 0.3229907751083374, | |
| "learning_rate": 0.0004951694816540476, | |
| "loss": 3.4337, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 8.774918490917559, | |
| "grad_norm": 0.3458191454410553, | |
| "learning_rate": 0.0004949947582993593, | |
| "loss": 3.4185, | |
| "step": 30150 | |
| }, | |
| { | |
| "epoch": 8.789473684210526, | |
| "grad_norm": 0.3303491175174713, | |
| "learning_rate": 0.0004948200349446709, | |
| "loss": 3.4228, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 8.804028877503493, | |
| "grad_norm": 0.32709747552871704, | |
| "learning_rate": 0.0004946453115899825, | |
| "loss": 3.4262, | |
| "step": 30250 | |
| }, | |
| { | |
| "epoch": 8.81858407079646, | |
| "grad_norm": 0.3337283134460449, | |
| "learning_rate": 0.000494470588235294, | |
| "loss": 3.4357, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 8.833139264089427, | |
| "grad_norm": 0.3414067327976227, | |
| "learning_rate": 0.0004942958648806057, | |
| "loss": 3.4318, | |
| "step": 30350 | |
| }, | |
| { | |
| "epoch": 8.847694457382394, | |
| "grad_norm": 0.32551151514053345, | |
| "learning_rate": 0.0004941211415259173, | |
| "loss": 3.421, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 8.862249650675361, | |
| "grad_norm": 0.35890769958496094, | |
| "learning_rate": 0.0004939464181712289, | |
| "loss": 3.4234, | |
| "step": 30450 | |
| }, | |
| { | |
| "epoch": 8.876804843968328, | |
| "grad_norm": 0.34184715151786804, | |
| "learning_rate": 0.0004937716948165404, | |
| "loss": 3.4403, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 8.891360037261295, | |
| "grad_norm": 0.35601386427879333, | |
| "learning_rate": 0.000493596971461852, | |
| "loss": 3.4282, | |
| "step": 30550 | |
| }, | |
| { | |
| "epoch": 8.905915230554262, | |
| "grad_norm": 0.3299999535083771, | |
| "learning_rate": 0.0004934222481071636, | |
| "loss": 3.4323, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 8.92047042384723, | |
| "grad_norm": 0.32472729682922363, | |
| "learning_rate": 0.0004932475247524751, | |
| "loss": 3.4419, | |
| "step": 30650 | |
| }, | |
| { | |
| "epoch": 8.935025617140196, | |
| "grad_norm": 0.3286599814891815, | |
| "learning_rate": 0.0004930728013977868, | |
| "loss": 3.4316, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 8.949580810433163, | |
| "grad_norm": 0.34092918038368225, | |
| "learning_rate": 0.0004928980780430984, | |
| "loss": 3.4298, | |
| "step": 30750 | |
| }, | |
| { | |
| "epoch": 8.964136003726129, | |
| "grad_norm": 0.3369322419166565, | |
| "learning_rate": 0.00049272335468841, | |
| "loss": 3.4254, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 8.978691197019096, | |
| "grad_norm": 0.3368324637413025, | |
| "learning_rate": 0.0004925486313337215, | |
| "loss": 3.4449, | |
| "step": 30850 | |
| }, | |
| { | |
| "epoch": 8.993246390312063, | |
| "grad_norm": 0.35143014788627625, | |
| "learning_rate": 0.0004923739079790332, | |
| "loss": 3.4482, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 9.007568700512342, | |
| "grad_norm": 0.3373304605484009, | |
| "learning_rate": 0.0004921991846243447, | |
| "loss": 3.3793, | |
| "step": 30950 | |
| }, | |
| { | |
| "epoch": 9.02212389380531, | |
| "grad_norm": 0.3433413505554199, | |
| "learning_rate": 0.0004920244612696563, | |
| "loss": 3.3271, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.02212389380531, | |
| "eval_accuracy": 0.3671869169543833, | |
| "eval_loss": 3.582354784011841, | |
| "eval_runtime": 82.4119, | |
| "eval_samples_per_second": 202.046, | |
| "eval_steps_per_second": 12.632, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.036679087098276, | |
| "grad_norm": 0.3288974165916443, | |
| "learning_rate": 0.0004918497379149679, | |
| "loss": 3.3322, | |
| "step": 31050 | |
| }, | |
| { | |
| "epoch": 9.051234280391244, | |
| "grad_norm": 0.33863672614097595, | |
| "learning_rate": 0.0004916750145602795, | |
| "loss": 3.336, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 9.06578947368421, | |
| "grad_norm": 0.3400559425354004, | |
| "learning_rate": 0.0004915002912055911, | |
| "loss": 3.3313, | |
| "step": 31150 | |
| }, | |
| { | |
| "epoch": 9.080344666977178, | |
| "grad_norm": 0.3689609169960022, | |
| "learning_rate": 0.0004913255678509026, | |
| "loss": 3.3425, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 9.094899860270145, | |
| "grad_norm": 0.3346099257469177, | |
| "learning_rate": 0.0004911508444962143, | |
| "loss": 3.3397, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 9.109455053563112, | |
| "grad_norm": 0.3258785605430603, | |
| "learning_rate": 0.0004909761211415259, | |
| "loss": 3.3493, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 9.124010246856079, | |
| "grad_norm": 0.3524854779243469, | |
| "learning_rate": 0.0004908013977868375, | |
| "loss": 3.3529, | |
| "step": 31350 | |
| }, | |
| { | |
| "epoch": 9.138565440149046, | |
| "grad_norm": 0.3655344843864441, | |
| "learning_rate": 0.0004906266744321491, | |
| "loss": 3.3435, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 9.153120633442011, | |
| "grad_norm": 0.3290146291255951, | |
| "learning_rate": 0.0004904519510774606, | |
| "loss": 3.3662, | |
| "step": 31450 | |
| }, | |
| { | |
| "epoch": 9.167675826734978, | |
| "grad_norm": 0.3558729887008667, | |
| "learning_rate": 0.0004902772277227722, | |
| "loss": 3.3503, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 9.182231020027945, | |
| "grad_norm": 0.34567442536354065, | |
| "learning_rate": 0.0004901025043680838, | |
| "loss": 3.3587, | |
| "step": 31550 | |
| }, | |
| { | |
| "epoch": 9.196786213320912, | |
| "grad_norm": 0.3287704885005951, | |
| "learning_rate": 0.0004899277810133955, | |
| "loss": 3.3608, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 9.21134140661388, | |
| "grad_norm": 0.3717208802700043, | |
| "learning_rate": 0.000489753057658707, | |
| "loss": 3.3679, | |
| "step": 31650 | |
| }, | |
| { | |
| "epoch": 9.225896599906847, | |
| "grad_norm": 0.32519346475601196, | |
| "learning_rate": 0.0004895783343040186, | |
| "loss": 3.3586, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 9.240451793199814, | |
| "grad_norm": 0.3226407766342163, | |
| "learning_rate": 0.0004894036109493302, | |
| "loss": 3.3655, | |
| "step": 31750 | |
| }, | |
| { | |
| "epoch": 9.25500698649278, | |
| "grad_norm": 0.3198281526565552, | |
| "learning_rate": 0.0004892288875946419, | |
| "loss": 3.3706, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 9.269562179785748, | |
| "grad_norm": 0.32234686613082886, | |
| "learning_rate": 0.0004890541642399534, | |
| "loss": 3.3704, | |
| "step": 31850 | |
| }, | |
| { | |
| "epoch": 9.284117373078715, | |
| "grad_norm": 0.3275204300880432, | |
| "learning_rate": 0.0004888794408852649, | |
| "loss": 3.3713, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 9.298672566371682, | |
| "grad_norm": 0.340788334608078, | |
| "learning_rate": 0.0004887047175305766, | |
| "loss": 3.3785, | |
| "step": 31950 | |
| }, | |
| { | |
| "epoch": 9.313227759664649, | |
| "grad_norm": 0.34868523478507996, | |
| "learning_rate": 0.0004885299941758881, | |
| "loss": 3.3778, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.313227759664649, | |
| "eval_accuracy": 0.3675521918196059, | |
| "eval_loss": 3.57724928855896, | |
| "eval_runtime": 82.3454, | |
| "eval_samples_per_second": 202.209, | |
| "eval_steps_per_second": 12.642, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.327782952957616, | |
| "grad_norm": 0.3213954269886017, | |
| "learning_rate": 0.0004883552708211997, | |
| "loss": 3.3795, | |
| "step": 32050 | |
| }, | |
| { | |
| "epoch": 9.342338146250583, | |
| "grad_norm": 0.34046775102615356, | |
| "learning_rate": 0.00048818054746651137, | |
| "loss": 3.3683, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 9.356893339543548, | |
| "grad_norm": 0.34967857599258423, | |
| "learning_rate": 0.0004880058241118229, | |
| "loss": 3.3822, | |
| "step": 32150 | |
| }, | |
| { | |
| "epoch": 9.371448532836515, | |
| "grad_norm": 0.3708345293998718, | |
| "learning_rate": 0.0004878311007571345, | |
| "loss": 3.3823, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 9.386003726129482, | |
| "grad_norm": 0.3245353400707245, | |
| "learning_rate": 0.0004876563774024461, | |
| "loss": 3.3809, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 9.40055891942245, | |
| "grad_norm": 0.3368681073188782, | |
| "learning_rate": 0.00048748165404775763, | |
| "loss": 3.3795, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 9.415114112715417, | |
| "grad_norm": 0.32234933972358704, | |
| "learning_rate": 0.0004873069306930693, | |
| "loss": 3.3903, | |
| "step": 32350 | |
| }, | |
| { | |
| "epoch": 9.429669306008384, | |
| "grad_norm": 0.32348915934562683, | |
| "learning_rate": 0.0004871322073383809, | |
| "loss": 3.3941, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 9.44422449930135, | |
| "grad_norm": 0.3842012882232666, | |
| "learning_rate": 0.00048695748398369247, | |
| "loss": 3.3732, | |
| "step": 32450 | |
| }, | |
| { | |
| "epoch": 9.458779692594318, | |
| "grad_norm": 0.3452703356742859, | |
| "learning_rate": 0.000486782760629004, | |
| "loss": 3.3864, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 9.473334885887285, | |
| "grad_norm": 0.3673316538333893, | |
| "learning_rate": 0.0004866080372743156, | |
| "loss": 3.379, | |
| "step": 32550 | |
| }, | |
| { | |
| "epoch": 9.487890079180252, | |
| "grad_norm": 0.35710281133651733, | |
| "learning_rate": 0.0004864333139196272, | |
| "loss": 3.3845, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 9.502445272473219, | |
| "grad_norm": 0.3384006917476654, | |
| "learning_rate": 0.00048625859056493885, | |
| "loss": 3.3826, | |
| "step": 32650 | |
| }, | |
| { | |
| "epoch": 9.517000465766186, | |
| "grad_norm": 0.3203440308570862, | |
| "learning_rate": 0.0004860838672102504, | |
| "loss": 3.3797, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 9.531555659059153, | |
| "grad_norm": 0.3406928479671478, | |
| "learning_rate": 0.000485909143855562, | |
| "loss": 3.3917, | |
| "step": 32750 | |
| }, | |
| { | |
| "epoch": 9.546110852352118, | |
| "grad_norm": 0.332173228263855, | |
| "learning_rate": 0.0004857344205008736, | |
| "loss": 3.392, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 9.560666045645085, | |
| "grad_norm": 0.333181768655777, | |
| "learning_rate": 0.00048555969714618517, | |
| "loss": 3.3943, | |
| "step": 32850 | |
| }, | |
| { | |
| "epoch": 9.575221238938052, | |
| "grad_norm": 0.32442259788513184, | |
| "learning_rate": 0.0004853849737914967, | |
| "loss": 3.402, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 9.58977643223102, | |
| "grad_norm": 0.34457752108573914, | |
| "learning_rate": 0.00048521025043680836, | |
| "loss": 3.395, | |
| "step": 32950 | |
| }, | |
| { | |
| "epoch": 9.604331625523987, | |
| "grad_norm": 0.3462928235530853, | |
| "learning_rate": 0.00048503552708211995, | |
| "loss": 3.3932, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.604331625523987, | |
| "eval_accuracy": 0.3678074611269623, | |
| "eval_loss": 3.573505401611328, | |
| "eval_runtime": 82.288, | |
| "eval_samples_per_second": 202.35, | |
| "eval_steps_per_second": 12.651, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.618886818816954, | |
| "grad_norm": 0.34157347679138184, | |
| "learning_rate": 0.00048486080372743155, | |
| "loss": 3.4026, | |
| "step": 33050 | |
| }, | |
| { | |
| "epoch": 9.63344201210992, | |
| "grad_norm": 0.32132917642593384, | |
| "learning_rate": 0.0004846860803727431, | |
| "loss": 3.4034, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 9.647997205402888, | |
| "grad_norm": 0.34497860074043274, | |
| "learning_rate": 0.0004845113570180547, | |
| "loss": 3.3994, | |
| "step": 33150 | |
| }, | |
| { | |
| "epoch": 9.662552398695855, | |
| "grad_norm": 0.3248150944709778, | |
| "learning_rate": 0.00048433663366336633, | |
| "loss": 3.3915, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 9.677107591988822, | |
| "grad_norm": 0.32516196370124817, | |
| "learning_rate": 0.0004841619103086779, | |
| "loss": 3.4019, | |
| "step": 33250 | |
| }, | |
| { | |
| "epoch": 9.691662785281789, | |
| "grad_norm": 0.3377247750759125, | |
| "learning_rate": 0.00048398718695398947, | |
| "loss": 3.4148, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 9.706217978574756, | |
| "grad_norm": 0.33178314566612244, | |
| "learning_rate": 0.00048381246359930106, | |
| "loss": 3.3975, | |
| "step": 33350 | |
| }, | |
| { | |
| "epoch": 9.720773171867723, | |
| "grad_norm": 0.36749276518821716, | |
| "learning_rate": 0.00048363774024461265, | |
| "loss": 3.4071, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 9.73532836516069, | |
| "grad_norm": 0.3314318358898163, | |
| "learning_rate": 0.0004834630168899242, | |
| "loss": 3.3958, | |
| "step": 33450 | |
| }, | |
| { | |
| "epoch": 9.749883558453657, | |
| "grad_norm": 0.6590105891227722, | |
| "learning_rate": 0.00048328829353523584, | |
| "loss": 3.3888, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 9.764438751746622, | |
| "grad_norm": 0.3279191255569458, | |
| "learning_rate": 0.00048311357018054744, | |
| "loss": 3.408, | |
| "step": 33550 | |
| }, | |
| { | |
| "epoch": 9.77899394503959, | |
| "grad_norm": 0.3082502484321594, | |
| "learning_rate": 0.00048293884682585903, | |
| "loss": 3.4074, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 9.793549138332557, | |
| "grad_norm": 0.32727038860321045, | |
| "learning_rate": 0.00048276412347117057, | |
| "loss": 3.4084, | |
| "step": 33650 | |
| }, | |
| { | |
| "epoch": 9.808104331625524, | |
| "grad_norm": 0.3583974540233612, | |
| "learning_rate": 0.00048258940011648217, | |
| "loss": 3.4103, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 9.82265952491849, | |
| "grad_norm": 0.35682162642478943, | |
| "learning_rate": 0.0004824146767617938, | |
| "loss": 3.4072, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 9.837214718211458, | |
| "grad_norm": 0.33762162923812866, | |
| "learning_rate": 0.0004822399534071054, | |
| "loss": 3.4168, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 9.851769911504425, | |
| "grad_norm": 0.32820042967796326, | |
| "learning_rate": 0.00048206523005241695, | |
| "loss": 3.4094, | |
| "step": 33850 | |
| }, | |
| { | |
| "epoch": 9.866325104797392, | |
| "grad_norm": 0.35341110825538635, | |
| "learning_rate": 0.00048189050669772854, | |
| "loss": 3.4201, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 9.880880298090359, | |
| "grad_norm": 0.33327147364616394, | |
| "learning_rate": 0.00048171578334304014, | |
| "loss": 3.4056, | |
| "step": 33950 | |
| }, | |
| { | |
| "epoch": 9.895435491383326, | |
| "grad_norm": 0.3687930703163147, | |
| "learning_rate": 0.00048154105998835173, | |
| "loss": 3.4236, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.895435491383326, | |
| "eval_accuracy": 0.36875003011637203, | |
| "eval_loss": 3.561856269836426, | |
| "eval_runtime": 82.4062, | |
| "eval_samples_per_second": 202.06, | |
| "eval_steps_per_second": 12.633, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.909990684676293, | |
| "grad_norm": 0.36275026202201843, | |
| "learning_rate": 0.0004813663366336633, | |
| "loss": 3.4118, | |
| "step": 34050 | |
| }, | |
| { | |
| "epoch": 9.92454587796926, | |
| "grad_norm": 0.32585081458091736, | |
| "learning_rate": 0.0004811916132789749, | |
| "loss": 3.408, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 9.939101071262227, | |
| "grad_norm": 0.37906891107559204, | |
| "learning_rate": 0.0004810168899242865, | |
| "loss": 3.4129, | |
| "step": 34150 | |
| }, | |
| { | |
| "epoch": 9.953656264555192, | |
| "grad_norm": 0.3377787172794342, | |
| "learning_rate": 0.0004808421665695981, | |
| "loss": 3.415, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 9.96821145784816, | |
| "grad_norm": 0.3153868019580841, | |
| "learning_rate": 0.00048066744321490965, | |
| "loss": 3.4114, | |
| "step": 34250 | |
| }, | |
| { | |
| "epoch": 9.982766651141127, | |
| "grad_norm": 0.3543281555175781, | |
| "learning_rate": 0.00048049271986022124, | |
| "loss": 3.4028, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 9.997321844434094, | |
| "grad_norm": 0.34016019105911255, | |
| "learning_rate": 0.0004803179965055329, | |
| "loss": 3.4015, | |
| "step": 34350 | |
| }, | |
| { | |
| "epoch": 10.011644154634373, | |
| "grad_norm": 0.34837228059768677, | |
| "learning_rate": 0.0004801432731508445, | |
| "loss": 3.3185, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 10.02619934792734, | |
| "grad_norm": 0.35867300629615784, | |
| "learning_rate": 0.000479968549796156, | |
| "loss": 3.3, | |
| "step": 34450 | |
| }, | |
| { | |
| "epoch": 10.040754541220307, | |
| "grad_norm": 0.35538730025291443, | |
| "learning_rate": 0.0004797938264414676, | |
| "loss": 3.3045, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 10.055309734513274, | |
| "grad_norm": 0.3725418746471405, | |
| "learning_rate": 0.0004796191030867792, | |
| "loss": 3.3033, | |
| "step": 34550 | |
| }, | |
| { | |
| "epoch": 10.069864927806242, | |
| "grad_norm": 0.3236791789531708, | |
| "learning_rate": 0.00047944437973209086, | |
| "loss": 3.3073, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 10.084420121099209, | |
| "grad_norm": 0.3377200961112976, | |
| "learning_rate": 0.0004792696563774024, | |
| "loss": 3.3133, | |
| "step": 34650 | |
| }, | |
| { | |
| "epoch": 10.098975314392176, | |
| "grad_norm": 0.3420577645301819, | |
| "learning_rate": 0.000479094933022714, | |
| "loss": 3.3183, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 10.113530507685143, | |
| "grad_norm": 0.34107503294944763, | |
| "learning_rate": 0.0004789202096680256, | |
| "loss": 3.299, | |
| "step": 34750 | |
| }, | |
| { | |
| "epoch": 10.12808570097811, | |
| "grad_norm": 0.3381388783454895, | |
| "learning_rate": 0.00047874548631333713, | |
| "loss": 3.3295, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 10.142640894271075, | |
| "grad_norm": 0.34578409790992737, | |
| "learning_rate": 0.0004785707629586487, | |
| "loss": 3.3355, | |
| "step": 34850 | |
| }, | |
| { | |
| "epoch": 10.157196087564042, | |
| "grad_norm": 0.36821314692497253, | |
| "learning_rate": 0.0004783960396039604, | |
| "loss": 3.3175, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 10.17175128085701, | |
| "grad_norm": 0.33993205428123474, | |
| "learning_rate": 0.00047822131624927197, | |
| "loss": 3.3359, | |
| "step": 34950 | |
| }, | |
| { | |
| "epoch": 10.186306474149976, | |
| "grad_norm": 0.3257310092449188, | |
| "learning_rate": 0.0004780465928945835, | |
| "loss": 3.3307, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.186306474149976, | |
| "eval_accuracy": 0.3683283421445513, | |
| "eval_loss": 3.5743184089660645, | |
| "eval_runtime": 82.3759, | |
| "eval_samples_per_second": 202.134, | |
| "eval_steps_per_second": 12.637, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.200861667442943, | |
| "grad_norm": 0.3305836617946625, | |
| "learning_rate": 0.0004778718695398951, | |
| "loss": 3.3578, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 10.21541686073591, | |
| "grad_norm": 0.36335787177085876, | |
| "learning_rate": 0.0004776971461852067, | |
| "loss": 3.3381, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 10.229972054028877, | |
| "grad_norm": 0.33552059531211853, | |
| "learning_rate": 0.00047752242283051835, | |
| "loss": 3.3348, | |
| "step": 35150 | |
| }, | |
| { | |
| "epoch": 10.244527247321844, | |
| "grad_norm": 0.35436898469924927, | |
| "learning_rate": 0.00047734769947582994, | |
| "loss": 3.3635, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 10.259082440614812, | |
| "grad_norm": 0.36603543162345886, | |
| "learning_rate": 0.0004771729761211415, | |
| "loss": 3.3415, | |
| "step": 35250 | |
| }, | |
| { | |
| "epoch": 10.273637633907779, | |
| "grad_norm": 0.3456842601299286, | |
| "learning_rate": 0.0004769982527664531, | |
| "loss": 3.3392, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 10.288192827200746, | |
| "grad_norm": 0.35775306820869446, | |
| "learning_rate": 0.00047682352941176467, | |
| "loss": 3.3642, | |
| "step": 35350 | |
| }, | |
| { | |
| "epoch": 10.302748020493713, | |
| "grad_norm": 0.3707828223705292, | |
| "learning_rate": 0.0004766488060570762, | |
| "loss": 3.3463, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 10.31730321378668, | |
| "grad_norm": 0.35976535081863403, | |
| "learning_rate": 0.00047647408270238786, | |
| "loss": 3.355, | |
| "step": 35450 | |
| }, | |
| { | |
| "epoch": 10.331858407079647, | |
| "grad_norm": 0.3616143465042114, | |
| "learning_rate": 0.00047629935934769945, | |
| "loss": 3.3493, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 10.346413600372612, | |
| "grad_norm": 0.3655940294265747, | |
| "learning_rate": 0.00047612463599301105, | |
| "loss": 3.3678, | |
| "step": 35550 | |
| }, | |
| { | |
| "epoch": 10.36096879366558, | |
| "grad_norm": 0.3340962529182434, | |
| "learning_rate": 0.0004759499126383226, | |
| "loss": 3.3547, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 10.375523986958546, | |
| "grad_norm": 0.36282655596733093, | |
| "learning_rate": 0.0004757751892836342, | |
| "loss": 3.3642, | |
| "step": 35650 | |
| }, | |
| { | |
| "epoch": 10.390079180251513, | |
| "grad_norm": 0.3513752222061157, | |
| "learning_rate": 0.0004756004659289458, | |
| "loss": 3.3572, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 10.40463437354448, | |
| "grad_norm": 0.3486036956310272, | |
| "learning_rate": 0.0004754257425742574, | |
| "loss": 3.3561, | |
| "step": 35750 | |
| }, | |
| { | |
| "epoch": 10.419189566837447, | |
| "grad_norm": 0.3694014847278595, | |
| "learning_rate": 0.00047525101921956896, | |
| "loss": 3.3611, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 10.433744760130415, | |
| "grad_norm": 0.36477726697921753, | |
| "learning_rate": 0.00047507629586488056, | |
| "loss": 3.3738, | |
| "step": 35850 | |
| }, | |
| { | |
| "epoch": 10.448299953423382, | |
| "grad_norm": 0.37289297580718994, | |
| "learning_rate": 0.00047490157251019215, | |
| "loss": 3.3576, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 10.462855146716349, | |
| "grad_norm": 0.3214474320411682, | |
| "learning_rate": 0.0004747268491555037, | |
| "loss": 3.3659, | |
| "step": 35950 | |
| }, | |
| { | |
| "epoch": 10.477410340009316, | |
| "grad_norm": 0.381782203912735, | |
| "learning_rate": 0.00047455212580081534, | |
| "loss": 3.3659, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.477410340009316, | |
| "eval_accuracy": 0.36878023463386306, | |
| "eval_loss": 3.567044734954834, | |
| "eval_runtime": 82.4838, | |
| "eval_samples_per_second": 201.87, | |
| "eval_steps_per_second": 12.621, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.491965533302283, | |
| "grad_norm": 0.35121598839759827, | |
| "learning_rate": 0.00047437740244612694, | |
| "loss": 3.3661, | |
| "step": 36050 | |
| }, | |
| { | |
| "epoch": 10.50652072659525, | |
| "grad_norm": 0.3433535695075989, | |
| "learning_rate": 0.00047420267909143853, | |
| "loss": 3.3642, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 10.521075919888217, | |
| "grad_norm": 0.3367346227169037, | |
| "learning_rate": 0.0004740279557367501, | |
| "loss": 3.3787, | |
| "step": 36150 | |
| }, | |
| { | |
| "epoch": 10.535631113181182, | |
| "grad_norm": 0.3348431885242462, | |
| "learning_rate": 0.00047385323238206166, | |
| "loss": 3.3748, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 10.55018630647415, | |
| "grad_norm": 0.3408311903476715, | |
| "learning_rate": 0.00047367850902737326, | |
| "loss": 3.3716, | |
| "step": 36250 | |
| }, | |
| { | |
| "epoch": 10.564741499767116, | |
| "grad_norm": 0.35345619916915894, | |
| "learning_rate": 0.0004735037856726849, | |
| "loss": 3.3793, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 10.579296693060083, | |
| "grad_norm": 0.3408675789833069, | |
| "learning_rate": 0.0004733290623179965, | |
| "loss": 3.3741, | |
| "step": 36350 | |
| }, | |
| { | |
| "epoch": 10.59385188635305, | |
| "grad_norm": 0.35938921570777893, | |
| "learning_rate": 0.00047315433896330804, | |
| "loss": 3.3737, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 10.608407079646017, | |
| "grad_norm": 0.33079761266708374, | |
| "learning_rate": 0.00047297961560861964, | |
| "loss": 3.378, | |
| "step": 36450 | |
| }, | |
| { | |
| "epoch": 10.622962272938985, | |
| "grad_norm": 0.34535646438598633, | |
| "learning_rate": 0.00047280489225393123, | |
| "loss": 3.3713, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 10.637517466231952, | |
| "grad_norm": 0.3378680646419525, | |
| "learning_rate": 0.0004726301688992429, | |
| "loss": 3.372, | |
| "step": 36550 | |
| }, | |
| { | |
| "epoch": 10.652072659524919, | |
| "grad_norm": 0.3342312276363373, | |
| "learning_rate": 0.0004724554455445544, | |
| "loss": 3.3845, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 10.666627852817886, | |
| "grad_norm": 0.33184394240379333, | |
| "learning_rate": 0.000472280722189866, | |
| "loss": 3.3807, | |
| "step": 36650 | |
| }, | |
| { | |
| "epoch": 10.681183046110853, | |
| "grad_norm": 0.33446264266967773, | |
| "learning_rate": 0.0004721059988351776, | |
| "loss": 3.3758, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 10.69573823940382, | |
| "grad_norm": 0.33704686164855957, | |
| "learning_rate": 0.00047193127548048915, | |
| "loss": 3.3844, | |
| "step": 36750 | |
| }, | |
| { | |
| "epoch": 10.710293432696787, | |
| "grad_norm": 0.33584555983543396, | |
| "learning_rate": 0.00047175655212580074, | |
| "loss": 3.384, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 10.724848625989754, | |
| "grad_norm": 0.3725135028362274, | |
| "learning_rate": 0.0004715818287711124, | |
| "loss": 3.3939, | |
| "step": 36850 | |
| }, | |
| { | |
| "epoch": 10.73940381928272, | |
| "grad_norm": 0.34769904613494873, | |
| "learning_rate": 0.000471407105416424, | |
| "loss": 3.3905, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 10.753959012575686, | |
| "grad_norm": 0.33186277747154236, | |
| "learning_rate": 0.0004712323820617355, | |
| "loss": 3.3899, | |
| "step": 36950 | |
| }, | |
| { | |
| "epoch": 10.768514205868653, | |
| "grad_norm": 0.35165971517562866, | |
| "learning_rate": 0.0004710576587070471, | |
| "loss": 3.3895, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.768514205868653, | |
| "eval_accuracy": 0.3692355354150318, | |
| "eval_loss": 3.561675548553467, | |
| "eval_runtime": 82.3387, | |
| "eval_samples_per_second": 202.226, | |
| "eval_steps_per_second": 12.643, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.78306939916162, | |
| "grad_norm": 0.32443472743034363, | |
| "learning_rate": 0.0004708829353523587, | |
| "loss": 3.3811, | |
| "step": 37050 | |
| }, | |
| { | |
| "epoch": 10.797624592454587, | |
| "grad_norm": 0.3643024265766144, | |
| "learning_rate": 0.0004707082119976703, | |
| "loss": 3.38, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 10.812179785747555, | |
| "grad_norm": 0.33836162090301514, | |
| "learning_rate": 0.0004705334886429819, | |
| "loss": 3.3942, | |
| "step": 37150 | |
| }, | |
| { | |
| "epoch": 10.826734979040522, | |
| "grad_norm": 0.3768630623817444, | |
| "learning_rate": 0.0004703587652882935, | |
| "loss": 3.3796, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 10.841290172333489, | |
| "grad_norm": 0.3157995343208313, | |
| "learning_rate": 0.0004701840419336051, | |
| "loss": 3.375, | |
| "step": 37250 | |
| }, | |
| { | |
| "epoch": 10.855845365626456, | |
| "grad_norm": 0.32104167342185974, | |
| "learning_rate": 0.0004700093185789167, | |
| "loss": 3.3902, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 10.870400558919423, | |
| "grad_norm": 0.3310007154941559, | |
| "learning_rate": 0.0004698345952242282, | |
| "loss": 3.3913, | |
| "step": 37350 | |
| }, | |
| { | |
| "epoch": 10.88495575221239, | |
| "grad_norm": 0.34951579570770264, | |
| "learning_rate": 0.00046965987186953987, | |
| "loss": 3.3937, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 10.899510945505357, | |
| "grad_norm": 0.37220627069473267, | |
| "learning_rate": 0.00046948514851485147, | |
| "loss": 3.3879, | |
| "step": 37450 | |
| }, | |
| { | |
| "epoch": 10.914066138798324, | |
| "grad_norm": 0.350894033908844, | |
| "learning_rate": 0.00046931042516016306, | |
| "loss": 3.3768, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 10.92862133209129, | |
| "grad_norm": 0.3326530158519745, | |
| "learning_rate": 0.0004691357018054746, | |
| "loss": 3.3807, | |
| "step": 37550 | |
| }, | |
| { | |
| "epoch": 10.943176525384256, | |
| "grad_norm": 0.35443612933158875, | |
| "learning_rate": 0.0004689609784507862, | |
| "loss": 3.3988, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 10.957731718677223, | |
| "grad_norm": 0.33639243245124817, | |
| "learning_rate": 0.0004687862550960978, | |
| "loss": 3.4011, | |
| "step": 37650 | |
| }, | |
| { | |
| "epoch": 10.97228691197019, | |
| "grad_norm": 0.3390151858329773, | |
| "learning_rate": 0.00046861153174140944, | |
| "loss": 3.3803, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 10.986842105263158, | |
| "grad_norm": 0.344147264957428, | |
| "learning_rate": 0.000468436808386721, | |
| "loss": 3.3828, | |
| "step": 37750 | |
| }, | |
| { | |
| "epoch": 11.001164415463437, | |
| "grad_norm": 0.36769360303878784, | |
| "learning_rate": 0.0004682620850320326, | |
| "loss": 3.3799, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 11.015719608756404, | |
| "grad_norm": 0.3531862199306488, | |
| "learning_rate": 0.00046808736167734417, | |
| "loss": 3.29, | |
| "step": 37850 | |
| }, | |
| { | |
| "epoch": 11.030274802049371, | |
| "grad_norm": 0.34574705362319946, | |
| "learning_rate": 0.0004679126383226557, | |
| "loss": 3.2814, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 11.044829995342338, | |
| "grad_norm": 0.3539099097251892, | |
| "learning_rate": 0.00046773791496796736, | |
| "loss": 3.2986, | |
| "step": 37950 | |
| }, | |
| { | |
| "epoch": 11.059385188635305, | |
| "grad_norm": 0.3517235517501831, | |
| "learning_rate": 0.00046756319161327895, | |
| "loss": 3.2783, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.059385188635305, | |
| "eval_accuracy": 0.36930804976247145, | |
| "eval_loss": 3.5699386596679688, | |
| "eval_runtime": 82.4781, | |
| "eval_samples_per_second": 201.884, | |
| "eval_steps_per_second": 12.622, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.073940381928272, | |
| "grad_norm": 0.3676995038986206, | |
| "learning_rate": 0.00046738846825859054, | |
| "loss": 3.3004, | |
| "step": 38050 | |
| }, | |
| { | |
| "epoch": 11.08849557522124, | |
| "grad_norm": 0.34967103600502014, | |
| "learning_rate": 0.0004672137449039021, | |
| "loss": 3.2921, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 11.103050768514207, | |
| "grad_norm": 0.3520808517932892, | |
| "learning_rate": 0.0004670390215492137, | |
| "loss": 3.2977, | |
| "step": 38150 | |
| }, | |
| { | |
| "epoch": 11.117605961807174, | |
| "grad_norm": 0.33675819635391235, | |
| "learning_rate": 0.0004668642981945253, | |
| "loss": 3.2804, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 11.132161155100139, | |
| "grad_norm": 0.3596917986869812, | |
| "learning_rate": 0.0004666895748398369, | |
| "loss": 3.2892, | |
| "step": 38250 | |
| }, | |
| { | |
| "epoch": 11.146716348393106, | |
| "grad_norm": 0.3737861216068268, | |
| "learning_rate": 0.00046651485148514846, | |
| "loss": 3.3037, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 11.161271541686073, | |
| "grad_norm": 0.34760457277297974, | |
| "learning_rate": 0.00046634012813046006, | |
| "loss": 3.3286, | |
| "step": 38350 | |
| }, | |
| { | |
| "epoch": 11.17582673497904, | |
| "grad_norm": 0.3329484760761261, | |
| "learning_rate": 0.00046616540477577165, | |
| "loss": 3.3199, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 11.190381928272007, | |
| "grad_norm": 0.35075056552886963, | |
| "learning_rate": 0.00046599068142108324, | |
| "loss": 3.3214, | |
| "step": 38450 | |
| }, | |
| { | |
| "epoch": 11.204937121564974, | |
| "grad_norm": 0.33573010563850403, | |
| "learning_rate": 0.0004658159580663948, | |
| "loss": 3.3132, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 11.219492314857941, | |
| "grad_norm": 0.3519783914089203, | |
| "learning_rate": 0.00046564123471170643, | |
| "loss": 3.3203, | |
| "step": 38550 | |
| }, | |
| { | |
| "epoch": 11.234047508150908, | |
| "grad_norm": 0.3524411618709564, | |
| "learning_rate": 0.00046546651135701803, | |
| "loss": 3.3205, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 11.248602701443875, | |
| "grad_norm": 0.3892838656902313, | |
| "learning_rate": 0.0004652917880023296, | |
| "loss": 3.3201, | |
| "step": 38650 | |
| }, | |
| { | |
| "epoch": 11.263157894736842, | |
| "grad_norm": 0.362114280462265, | |
| "learning_rate": 0.00046511706464764116, | |
| "loss": 3.3319, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 11.27771308802981, | |
| "grad_norm": 0.3446020781993866, | |
| "learning_rate": 0.00046494234129295276, | |
| "loss": 3.3307, | |
| "step": 38750 | |
| }, | |
| { | |
| "epoch": 11.292268281322777, | |
| "grad_norm": 0.3528365194797516, | |
| "learning_rate": 0.0004647676179382644, | |
| "loss": 3.3245, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 11.306823474615744, | |
| "grad_norm": 0.3371918201446533, | |
| "learning_rate": 0.000464592894583576, | |
| "loss": 3.3284, | |
| "step": 38850 | |
| }, | |
| { | |
| "epoch": 11.32137866790871, | |
| "grad_norm": 0.3414005935192108, | |
| "learning_rate": 0.00046441817122888754, | |
| "loss": 3.3329, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 11.335933861201676, | |
| "grad_norm": 0.32874834537506104, | |
| "learning_rate": 0.00046424344787419913, | |
| "loss": 3.3439, | |
| "step": 38950 | |
| }, | |
| { | |
| "epoch": 11.350489054494643, | |
| "grad_norm": 0.37445372343063354, | |
| "learning_rate": 0.00046406872451951073, | |
| "loss": 3.3433, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.350489054494643, | |
| "eval_accuracy": 0.3693060517982794, | |
| "eval_loss": 3.5668535232543945, | |
| "eval_runtime": 82.3281, | |
| "eval_samples_per_second": 202.252, | |
| "eval_steps_per_second": 12.645, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.36504424778761, | |
| "grad_norm": 0.35784077644348145, | |
| "learning_rate": 0.00046389400116482227, | |
| "loss": 3.3307, | |
| "step": 39050 | |
| }, | |
| { | |
| "epoch": 11.379599441080577, | |
| "grad_norm": 0.33896124362945557, | |
| "learning_rate": 0.0004637192778101339, | |
| "loss": 3.3395, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 11.394154634373544, | |
| "grad_norm": 0.3631598651409149, | |
| "learning_rate": 0.0004635445544554455, | |
| "loss": 3.3318, | |
| "step": 39150 | |
| }, | |
| { | |
| "epoch": 11.408709827666511, | |
| "grad_norm": 0.3224392831325531, | |
| "learning_rate": 0.0004633698311007571, | |
| "loss": 3.3472, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 11.423265020959478, | |
| "grad_norm": 0.3684169054031372, | |
| "learning_rate": 0.0004631951077460687, | |
| "loss": 3.3294, | |
| "step": 39250 | |
| }, | |
| { | |
| "epoch": 11.437820214252445, | |
| "grad_norm": 0.34574300050735474, | |
| "learning_rate": 0.00046302038439138024, | |
| "loss": 3.3378, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 11.452375407545413, | |
| "grad_norm": 0.34315943717956543, | |
| "learning_rate": 0.0004628456610366919, | |
| "loss": 3.3433, | |
| "step": 39350 | |
| }, | |
| { | |
| "epoch": 11.46693060083838, | |
| "grad_norm": 0.32054659724235535, | |
| "learning_rate": 0.0004626709376820035, | |
| "loss": 3.3426, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 11.481485794131347, | |
| "grad_norm": 0.3594513535499573, | |
| "learning_rate": 0.0004624962143273151, | |
| "loss": 3.352, | |
| "step": 39450 | |
| }, | |
| { | |
| "epoch": 11.496040987424314, | |
| "grad_norm": 0.33606380224227905, | |
| "learning_rate": 0.0004623214909726266, | |
| "loss": 3.3349, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 11.51059618071728, | |
| "grad_norm": 0.36756107211112976, | |
| "learning_rate": 0.0004621467676179382, | |
| "loss": 3.3543, | |
| "step": 39550 | |
| }, | |
| { | |
| "epoch": 11.525151374010246, | |
| "grad_norm": 0.3974190056324005, | |
| "learning_rate": 0.0004619720442632498, | |
| "loss": 3.3607, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 11.539706567303213, | |
| "grad_norm": 0.3304991126060486, | |
| "learning_rate": 0.00046179732090856145, | |
| "loss": 3.3499, | |
| "step": 39650 | |
| }, | |
| { | |
| "epoch": 11.55426176059618, | |
| "grad_norm": 0.32750317454338074, | |
| "learning_rate": 0.000461622597553873, | |
| "loss": 3.3465, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 11.568816953889147, | |
| "grad_norm": 0.3407932221889496, | |
| "learning_rate": 0.0004614478741991846, | |
| "loss": 3.3571, | |
| "step": 39750 | |
| }, | |
| { | |
| "epoch": 11.583372147182114, | |
| "grad_norm": 0.34489479660987854, | |
| "learning_rate": 0.0004612731508444962, | |
| "loss": 3.3538, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 11.597927340475081, | |
| "grad_norm": 0.3526317775249481, | |
| "learning_rate": 0.0004610984274898077, | |
| "loss": 3.353, | |
| "step": 39850 | |
| }, | |
| { | |
| "epoch": 11.612482533768048, | |
| "grad_norm": 0.34726589918136597, | |
| "learning_rate": 0.00046092370413511937, | |
| "loss": 3.3611, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 11.627037727061015, | |
| "grad_norm": 0.38080501556396484, | |
| "learning_rate": 0.00046074898078043096, | |
| "loss": 3.3543, | |
| "step": 39950 | |
| }, | |
| { | |
| "epoch": 11.641592920353983, | |
| "grad_norm": 0.3480202257633209, | |
| "learning_rate": 0.00046057425742574256, | |
| "loss": 3.3626, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.641592920353983, | |
| "eval_accuracy": 0.3696809639025459, | |
| "eval_loss": 3.5595285892486572, | |
| "eval_runtime": 82.3967, | |
| "eval_samples_per_second": 202.083, | |
| "eval_steps_per_second": 12.634, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.65614811364695, | |
| "grad_norm": 0.34867966175079346, | |
| "learning_rate": 0.0004603995340710541, | |
| "loss": 3.3462, | |
| "step": 40050 | |
| }, | |
| { | |
| "epoch": 11.670703306939917, | |
| "grad_norm": 0.3449127972126007, | |
| "learning_rate": 0.0004602248107163657, | |
| "loss": 3.373, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 11.685258500232884, | |
| "grad_norm": 0.34967654943466187, | |
| "learning_rate": 0.0004600500873616773, | |
| "loss": 3.3622, | |
| "step": 40150 | |
| }, | |
| { | |
| "epoch": 11.69981369352585, | |
| "grad_norm": 0.33432042598724365, | |
| "learning_rate": 0.00045987536400698894, | |
| "loss": 3.3664, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 11.714368886818818, | |
| "grad_norm": 0.32959768176078796, | |
| "learning_rate": 0.0004597006406523005, | |
| "loss": 3.3685, | |
| "step": 40250 | |
| }, | |
| { | |
| "epoch": 11.728924080111783, | |
| "grad_norm": 0.3566364645957947, | |
| "learning_rate": 0.00045952591729761207, | |
| "loss": 3.3612, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 11.74347927340475, | |
| "grad_norm": 0.3422619700431824, | |
| "learning_rate": 0.00045935119394292367, | |
| "loss": 3.363, | |
| "step": 40350 | |
| }, | |
| { | |
| "epoch": 11.758034466697717, | |
| "grad_norm": 0.3390800952911377, | |
| "learning_rate": 0.00045917647058823526, | |
| "loss": 3.3654, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 11.772589659990684, | |
| "grad_norm": 0.3512347638607025, | |
| "learning_rate": 0.0004590017472335468, | |
| "loss": 3.3583, | |
| "step": 40450 | |
| }, | |
| { | |
| "epoch": 11.787144853283651, | |
| "grad_norm": 0.40144476294517517, | |
| "learning_rate": 0.00045882702387885845, | |
| "loss": 3.3674, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 11.801700046576618, | |
| "grad_norm": 0.3194873332977295, | |
| "learning_rate": 0.00045865230052417004, | |
| "loss": 3.3649, | |
| "step": 40550 | |
| }, | |
| { | |
| "epoch": 11.816255239869585, | |
| "grad_norm": 0.33631259202957153, | |
| "learning_rate": 0.00045847757716948164, | |
| "loss": 3.3815, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 11.830810433162553, | |
| "grad_norm": 0.3417883813381195, | |
| "learning_rate": 0.0004583028538147932, | |
| "loss": 3.3607, | |
| "step": 40650 | |
| }, | |
| { | |
| "epoch": 11.84536562645552, | |
| "grad_norm": 0.33968234062194824, | |
| "learning_rate": 0.00045812813046010477, | |
| "loss": 3.3733, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 11.859920819748487, | |
| "grad_norm": 0.3497338593006134, | |
| "learning_rate": 0.0004579534071054164, | |
| "loss": 3.3688, | |
| "step": 40750 | |
| }, | |
| { | |
| "epoch": 11.874476013041454, | |
| "grad_norm": 0.33216750621795654, | |
| "learning_rate": 0.000457778683750728, | |
| "loss": 3.3823, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 11.88903120633442, | |
| "grad_norm": 0.3574502170085907, | |
| "learning_rate": 0.00045760396039603955, | |
| "loss": 3.3729, | |
| "step": 40850 | |
| }, | |
| { | |
| "epoch": 11.903586399627388, | |
| "grad_norm": 0.35262975096702576, | |
| "learning_rate": 0.00045742923704135115, | |
| "loss": 3.3734, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 11.918141592920353, | |
| "grad_norm": 0.34812983870506287, | |
| "learning_rate": 0.00045725451368666274, | |
| "loss": 3.3727, | |
| "step": 40950 | |
| }, | |
| { | |
| "epoch": 11.93269678621332, | |
| "grad_norm": 0.3207479417324066, | |
| "learning_rate": 0.0004570797903319743, | |
| "loss": 3.3781, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.93269678621332, | |
| "eval_accuracy": 0.37058651179075064, | |
| "eval_loss": 3.547286033630371, | |
| "eval_runtime": 82.3186, | |
| "eval_samples_per_second": 202.275, | |
| "eval_steps_per_second": 12.646, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.947251979506287, | |
| "grad_norm": 0.33297690749168396, | |
| "learning_rate": 0.00045690506697728593, | |
| "loss": 3.3742, | |
| "step": 41050 | |
| }, | |
| { | |
| "epoch": 11.961807172799254, | |
| "grad_norm": 0.3312673270702362, | |
| "learning_rate": 0.0004567303436225975, | |
| "loss": 3.3701, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 11.976362366092221, | |
| "grad_norm": 0.3394164741039276, | |
| "learning_rate": 0.0004565556202679091, | |
| "loss": 3.3702, | |
| "step": 41150 | |
| }, | |
| { | |
| "epoch": 11.990917559385188, | |
| "grad_norm": 0.3303782343864441, | |
| "learning_rate": 0.00045638089691322066, | |
| "loss": 3.3663, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 12.005239869585468, | |
| "grad_norm": 0.3281562924385071, | |
| "learning_rate": 0.00045620617355853225, | |
| "loss": 3.3378, | |
| "step": 41250 | |
| }, | |
| { | |
| "epoch": 12.019795062878435, | |
| "grad_norm": 0.3638324737548828, | |
| "learning_rate": 0.0004560314502038439, | |
| "loss": 3.2617, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 12.034350256171402, | |
| "grad_norm": 0.35445764660835266, | |
| "learning_rate": 0.0004558567268491555, | |
| "loss": 3.264, | |
| "step": 41350 | |
| }, | |
| { | |
| "epoch": 12.04890544946437, | |
| "grad_norm": 0.3631881773471832, | |
| "learning_rate": 0.00045568200349446704, | |
| "loss": 3.264, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 12.063460642757336, | |
| "grad_norm": 0.34890100359916687, | |
| "learning_rate": 0.00045550728013977863, | |
| "loss": 3.2642, | |
| "step": 41450 | |
| }, | |
| { | |
| "epoch": 12.078015836050303, | |
| "grad_norm": 0.34587299823760986, | |
| "learning_rate": 0.0004553325567850902, | |
| "loss": 3.284, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 12.09257102934327, | |
| "grad_norm": 0.3530641198158264, | |
| "learning_rate": 0.0004551578334304018, | |
| "loss": 3.2813, | |
| "step": 41550 | |
| }, | |
| { | |
| "epoch": 12.107126222636236, | |
| "grad_norm": 0.36315974593162537, | |
| "learning_rate": 0.00045498311007571347, | |
| "loss": 3.2905, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 12.121681415929203, | |
| "grad_norm": 0.3360583186149597, | |
| "learning_rate": 0.000454808386721025, | |
| "loss": 3.2944, | |
| "step": 41650 | |
| }, | |
| { | |
| "epoch": 12.13623660922217, | |
| "grad_norm": 0.340507447719574, | |
| "learning_rate": 0.0004546336633663366, | |
| "loss": 3.2843, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 12.150791802515137, | |
| "grad_norm": 0.3421109616756439, | |
| "learning_rate": 0.0004544589400116482, | |
| "loss": 3.2895, | |
| "step": 41750 | |
| }, | |
| { | |
| "epoch": 12.165346995808104, | |
| "grad_norm": 0.34345632791519165, | |
| "learning_rate": 0.00045428421665695974, | |
| "loss": 3.2925, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 12.179902189101071, | |
| "grad_norm": 0.3459620773792267, | |
| "learning_rate": 0.00045410949330227133, | |
| "loss": 3.3009, | |
| "step": 41850 | |
| }, | |
| { | |
| "epoch": 12.194457382394038, | |
| "grad_norm": 0.36432555317878723, | |
| "learning_rate": 0.000453934769947583, | |
| "loss": 3.3021, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 12.209012575687005, | |
| "grad_norm": 0.3652282655239105, | |
| "learning_rate": 0.0004537600465928946, | |
| "loss": 3.3033, | |
| "step": 41950 | |
| }, | |
| { | |
| "epoch": 12.223567768979972, | |
| "grad_norm": 0.36687561869621277, | |
| "learning_rate": 0.0004535853232382061, | |
| "loss": 3.3113, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.223567768979972, | |
| "eval_accuracy": 0.3698935707980374, | |
| "eval_loss": 3.561998128890991, | |
| "eval_runtime": 82.3887, | |
| "eval_samples_per_second": 202.103, | |
| "eval_steps_per_second": 12.635, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.23812296227294, | |
| "grad_norm": 0.3396989703178406, | |
| "learning_rate": 0.0004534105998835177, | |
| "loss": 3.3021, | |
| "step": 42050 | |
| }, | |
| { | |
| "epoch": 12.252678155565906, | |
| "grad_norm": 0.35155773162841797, | |
| "learning_rate": 0.0004532358765288293, | |
| "loss": 3.2982, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 12.267233348858873, | |
| "grad_norm": 0.3596819043159485, | |
| "learning_rate": 0.00045306115317414095, | |
| "loss": 3.3086, | |
| "step": 42150 | |
| }, | |
| { | |
| "epoch": 12.28178854215184, | |
| "grad_norm": 0.37626877427101135, | |
| "learning_rate": 0.0004528864298194525, | |
| "loss": 3.316, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 12.296343735444808, | |
| "grad_norm": 0.3309476375579834, | |
| "learning_rate": 0.0004527117064647641, | |
| "loss": 3.3061, | |
| "step": 42250 | |
| }, | |
| { | |
| "epoch": 12.310898928737773, | |
| "grad_norm": 0.3493840992450714, | |
| "learning_rate": 0.0004525369831100757, | |
| "loss": 3.3012, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 12.32545412203074, | |
| "grad_norm": 0.35301321744918823, | |
| "learning_rate": 0.0004523622597553872, | |
| "loss": 3.3314, | |
| "step": 42350 | |
| }, | |
| { | |
| "epoch": 12.340009315323707, | |
| "grad_norm": 0.3557305335998535, | |
| "learning_rate": 0.0004521875364006988, | |
| "loss": 3.3129, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 12.354564508616674, | |
| "grad_norm": 0.383173406124115, | |
| "learning_rate": 0.00045201281304601046, | |
| "loss": 3.3183, | |
| "step": 42450 | |
| }, | |
| { | |
| "epoch": 12.369119701909641, | |
| "grad_norm": 0.3709447681903839, | |
| "learning_rate": 0.00045183808969132206, | |
| "loss": 3.3179, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 12.383674895202608, | |
| "grad_norm": 0.33501526713371277, | |
| "learning_rate": 0.00045166336633663365, | |
| "loss": 3.3281, | |
| "step": 42550 | |
| }, | |
| { | |
| "epoch": 12.398230088495575, | |
| "grad_norm": 0.3576209247112274, | |
| "learning_rate": 0.0004514886429819452, | |
| "loss": 3.3149, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 12.412785281788542, | |
| "grad_norm": 0.34453296661376953, | |
| "learning_rate": 0.0004513139196272568, | |
| "loss": 3.3268, | |
| "step": 42650 | |
| }, | |
| { | |
| "epoch": 12.42734047508151, | |
| "grad_norm": 0.36603885889053345, | |
| "learning_rate": 0.00045113919627256843, | |
| "loss": 3.3169, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 12.441895668374476, | |
| "grad_norm": 0.3940148651599884, | |
| "learning_rate": 0.00045096447291788003, | |
| "loss": 3.3225, | |
| "step": 42750 | |
| }, | |
| { | |
| "epoch": 12.456450861667443, | |
| "grad_norm": 0.32357311248779297, | |
| "learning_rate": 0.00045078974956319157, | |
| "loss": 3.3327, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 12.47100605496041, | |
| "grad_norm": 0.3466281294822693, | |
| "learning_rate": 0.00045061502620850316, | |
| "loss": 3.3226, | |
| "step": 42850 | |
| }, | |
| { | |
| "epoch": 12.485561248253378, | |
| "grad_norm": 0.36143070459365845, | |
| "learning_rate": 0.00045044030285381476, | |
| "loss": 3.3289, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 12.500116441546343, | |
| "grad_norm": 0.33683663606643677, | |
| "learning_rate": 0.0004502655794991263, | |
| "loss": 3.3292, | |
| "step": 42950 | |
| }, | |
| { | |
| "epoch": 12.51467163483931, | |
| "grad_norm": 0.33424046635627747, | |
| "learning_rate": 0.00045009085614443795, | |
| "loss": 3.3312, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.51467163483931, | |
| "eval_accuracy": 0.3704625204835402, | |
| "eval_loss": 3.5558414459228516, | |
| "eval_runtime": 82.3517, | |
| "eval_samples_per_second": 202.194, | |
| "eval_steps_per_second": 12.641, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.529226828132277, | |
| "grad_norm": 0.348614901304245, | |
| "learning_rate": 0.00044991613278974954, | |
| "loss": 3.3363, | |
| "step": 43050 | |
| }, | |
| { | |
| "epoch": 12.543782021425244, | |
| "grad_norm": 0.3651944398880005, | |
| "learning_rate": 0.00044974140943506113, | |
| "loss": 3.3379, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 12.558337214718211, | |
| "grad_norm": 0.36780694127082825, | |
| "learning_rate": 0.0004495666860803727, | |
| "loss": 3.3291, | |
| "step": 43150 | |
| }, | |
| { | |
| "epoch": 12.572892408011178, | |
| "grad_norm": 0.36323991417884827, | |
| "learning_rate": 0.00044939196272568427, | |
| "loss": 3.3362, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 12.587447601304145, | |
| "grad_norm": 0.35772693157196045, | |
| "learning_rate": 0.00044921723937099586, | |
| "loss": 3.3296, | |
| "step": 43250 | |
| }, | |
| { | |
| "epoch": 12.602002794597112, | |
| "grad_norm": 0.35034236311912537, | |
| "learning_rate": 0.0004490425160163075, | |
| "loss": 3.3504, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 12.61655798789008, | |
| "grad_norm": 0.38235005736351013, | |
| "learning_rate": 0.00044886779266161905, | |
| "loss": 3.3308, | |
| "step": 43350 | |
| }, | |
| { | |
| "epoch": 12.631113181183046, | |
| "grad_norm": 0.3552211821079254, | |
| "learning_rate": 0.00044869306930693065, | |
| "loss": 3.3436, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 12.645668374476013, | |
| "grad_norm": 0.3449898064136505, | |
| "learning_rate": 0.00044851834595224224, | |
| "loss": 3.3445, | |
| "step": 43450 | |
| }, | |
| { | |
| "epoch": 12.66022356776898, | |
| "grad_norm": 0.3679625988006592, | |
| "learning_rate": 0.00044834362259755383, | |
| "loss": 3.3472, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 12.674778761061948, | |
| "grad_norm": 0.36120590567588806, | |
| "learning_rate": 0.00044816889924286543, | |
| "loss": 3.3389, | |
| "step": 43550 | |
| }, | |
| { | |
| "epoch": 12.689333954354915, | |
| "grad_norm": 0.3396338224411011, | |
| "learning_rate": 0.000447994175888177, | |
| "loss": 3.3461, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 12.703889147647882, | |
| "grad_norm": 0.3563362658023834, | |
| "learning_rate": 0.0004478194525334886, | |
| "loss": 3.3512, | |
| "step": 43650 | |
| }, | |
| { | |
| "epoch": 12.718444340940847, | |
| "grad_norm": 0.3691624402999878, | |
| "learning_rate": 0.0004476447291788002, | |
| "loss": 3.3385, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 12.732999534233814, | |
| "grad_norm": 0.3304498791694641, | |
| "learning_rate": 0.00044747000582411175, | |
| "loss": 3.3383, | |
| "step": 43750 | |
| }, | |
| { | |
| "epoch": 12.747554727526781, | |
| "grad_norm": 0.34523892402648926, | |
| "learning_rate": 0.00044729528246942335, | |
| "loss": 3.3408, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 12.762109920819748, | |
| "grad_norm": 0.31643256545066833, | |
| "learning_rate": 0.000447120559114735, | |
| "loss": 3.3326, | |
| "step": 43850 | |
| }, | |
| { | |
| "epoch": 12.776665114112715, | |
| "grad_norm": 0.34798404574394226, | |
| "learning_rate": 0.0004469458357600466, | |
| "loss": 3.3465, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 12.791220307405682, | |
| "grad_norm": 0.35351231694221497, | |
| "learning_rate": 0.00044677111240535813, | |
| "loss": 3.3471, | |
| "step": 43950 | |
| }, | |
| { | |
| "epoch": 12.80577550069865, | |
| "grad_norm": 0.34743833541870117, | |
| "learning_rate": 0.0004465963890506697, | |
| "loss": 3.3535, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.80577550069865, | |
| "eval_accuracy": 0.37095648774818973, | |
| "eval_loss": 3.5515782833099365, | |
| "eval_runtime": 82.4089, | |
| "eval_samples_per_second": 202.053, | |
| "eval_steps_per_second": 12.632, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.820330693991616, | |
| "grad_norm": 0.32085710763931274, | |
| "learning_rate": 0.0004464216656959813, | |
| "loss": 3.3588, | |
| "step": 44050 | |
| }, | |
| { | |
| "epoch": 12.834885887284583, | |
| "grad_norm": 0.34186598658561707, | |
| "learning_rate": 0.00044624694234129297, | |
| "loss": 3.3603, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 12.84944108057755, | |
| "grad_norm": 0.3461160659790039, | |
| "learning_rate": 0.0004460722189866045, | |
| "loss": 3.3494, | |
| "step": 44150 | |
| }, | |
| { | |
| "epoch": 12.863996273870518, | |
| "grad_norm": 0.3502444922924042, | |
| "learning_rate": 0.0004458974956319161, | |
| "loss": 3.3514, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 12.878551467163485, | |
| "grad_norm": 0.3411875069141388, | |
| "learning_rate": 0.0004457227722772277, | |
| "loss": 3.3518, | |
| "step": 44250 | |
| }, | |
| { | |
| "epoch": 12.89310666045645, | |
| "grad_norm": 0.3508564233779907, | |
| "learning_rate": 0.00044554804892253923, | |
| "loss": 3.3507, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 12.907661853749417, | |
| "grad_norm": 0.3484634757041931, | |
| "learning_rate": 0.00044537332556785083, | |
| "loss": 3.3491, | |
| "step": 44350 | |
| }, | |
| { | |
| "epoch": 12.922217047042384, | |
| "grad_norm": 0.360052227973938, | |
| "learning_rate": 0.0004451986022131625, | |
| "loss": 3.3606, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 12.936772240335351, | |
| "grad_norm": 0.33051279187202454, | |
| "learning_rate": 0.00044502387885847407, | |
| "loss": 3.3551, | |
| "step": 44450 | |
| }, | |
| { | |
| "epoch": 12.951327433628318, | |
| "grad_norm": 0.3616723120212555, | |
| "learning_rate": 0.0004448491555037856, | |
| "loss": 3.3526, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 12.965882626921285, | |
| "grad_norm": 0.3670650124549866, | |
| "learning_rate": 0.0004446744321490972, | |
| "loss": 3.3509, | |
| "step": 44550 | |
| }, | |
| { | |
| "epoch": 12.980437820214252, | |
| "grad_norm": 0.3474695086479187, | |
| "learning_rate": 0.0004444997087944088, | |
| "loss": 3.361, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 12.99499301350722, | |
| "grad_norm": 0.3489047884941101, | |
| "learning_rate": 0.0004443249854397204, | |
| "loss": 3.349, | |
| "step": 44650 | |
| }, | |
| { | |
| "epoch": 13.009315323707499, | |
| "grad_norm": 0.3696801960468292, | |
| "learning_rate": 0.000444150262085032, | |
| "loss": 3.2872, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 13.023870517000466, | |
| "grad_norm": 0.376399427652359, | |
| "learning_rate": 0.0004439755387303436, | |
| "loss": 3.248, | |
| "step": 44750 | |
| }, | |
| { | |
| "epoch": 13.038425710293433, | |
| "grad_norm": 0.3511487543582916, | |
| "learning_rate": 0.0004438008153756552, | |
| "loss": 3.2532, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 13.0529809035864, | |
| "grad_norm": 0.3696863353252411, | |
| "learning_rate": 0.00044362609202096677, | |
| "loss": 3.2593, | |
| "step": 44850 | |
| }, | |
| { | |
| "epoch": 13.067536096879367, | |
| "grad_norm": 0.35395491123199463, | |
| "learning_rate": 0.0004434513686662783, | |
| "loss": 3.2472, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 13.082091290172334, | |
| "grad_norm": 0.35846418142318726, | |
| "learning_rate": 0.00044327664531158996, | |
| "loss": 3.2505, | |
| "step": 44950 | |
| }, | |
| { | |
| "epoch": 13.0966464834653, | |
| "grad_norm": 0.3505629301071167, | |
| "learning_rate": 0.00044310192195690155, | |
| "loss": 3.2658, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.0966464834653, | |
| "eval_accuracy": 0.3705427916331371, | |
| "eval_loss": 3.5597288608551025, | |
| "eval_runtime": 82.3291, | |
| "eval_samples_per_second": 202.249, | |
| "eval_steps_per_second": 12.644, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.111201676758267, | |
| "grad_norm": 0.3679329752922058, | |
| "learning_rate": 0.00044292719860221315, | |
| "loss": 3.2686, | |
| "step": 45050 | |
| }, | |
| { | |
| "epoch": 13.125756870051234, | |
| "grad_norm": 0.376454621553421, | |
| "learning_rate": 0.0004427524752475247, | |
| "loss": 3.2631, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 13.1403120633442, | |
| "grad_norm": 0.3737456500530243, | |
| "learning_rate": 0.0004425777518928363, | |
| "loss": 3.2665, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 13.154867256637168, | |
| "grad_norm": 0.37014853954315186, | |
| "learning_rate": 0.0004424030285381479, | |
| "loss": 3.2785, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 13.169422449930135, | |
| "grad_norm": 0.37406471371650696, | |
| "learning_rate": 0.0004422283051834595, | |
| "loss": 3.2711, | |
| "step": 45250 | |
| }, | |
| { | |
| "epoch": 13.183977643223102, | |
| "grad_norm": 0.3682869076728821, | |
| "learning_rate": 0.00044205358182877107, | |
| "loss": 3.2935, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 13.198532836516069, | |
| "grad_norm": 0.3566969335079193, | |
| "learning_rate": 0.00044187885847408266, | |
| "loss": 3.293, | |
| "step": 45350 | |
| }, | |
| { | |
| "epoch": 13.213088029809036, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00044170413511939425, | |
| "loss": 3.2909, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 13.227643223102003, | |
| "grad_norm": 0.3529301881790161, | |
| "learning_rate": 0.0004415294117647058, | |
| "loss": 3.2907, | |
| "step": 45450 | |
| }, | |
| { | |
| "epoch": 13.24219841639497, | |
| "grad_norm": 0.32464173436164856, | |
| "learning_rate": 0.00044135468841001744, | |
| "loss": 3.2965, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 13.256753609687937, | |
| "grad_norm": 0.35889777541160583, | |
| "learning_rate": 0.00044117996505532904, | |
| "loss": 3.2841, | |
| "step": 45550 | |
| }, | |
| { | |
| "epoch": 13.271308802980904, | |
| "grad_norm": 0.35210585594177246, | |
| "learning_rate": 0.00044100524170064063, | |
| "loss": 3.2876, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 13.285863996273871, | |
| "grad_norm": 0.34717512130737305, | |
| "learning_rate": 0.0004408305183459522, | |
| "loss": 3.2919, | |
| "step": 45650 | |
| }, | |
| { | |
| "epoch": 13.300419189566837, | |
| "grad_norm": 0.3792016804218292, | |
| "learning_rate": 0.00044065579499126377, | |
| "loss": 3.2972, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 13.314974382859804, | |
| "grad_norm": 0.3483853042125702, | |
| "learning_rate": 0.00044048107163657536, | |
| "loss": 3.308, | |
| "step": 45750 | |
| }, | |
| { | |
| "epoch": 13.32952957615277, | |
| "grad_norm": 0.3516431152820587, | |
| "learning_rate": 0.000440306348281887, | |
| "loss": 3.3038, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 13.344084769445738, | |
| "grad_norm": 0.3443547785282135, | |
| "learning_rate": 0.0004401316249271986, | |
| "loss": 3.3073, | |
| "step": 45850 | |
| }, | |
| { | |
| "epoch": 13.358639962738705, | |
| "grad_norm": 0.3471525311470032, | |
| "learning_rate": 0.00043995690157251014, | |
| "loss": 3.3026, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 13.373195156031672, | |
| "grad_norm": 0.3602656424045563, | |
| "learning_rate": 0.00043978217821782174, | |
| "loss": 3.3115, | |
| "step": 45950 | |
| }, | |
| { | |
| "epoch": 13.387750349324639, | |
| "grad_norm": 0.36995962262153625, | |
| "learning_rate": 0.00043960745486313333, | |
| "loss": 3.3048, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.387750349324639, | |
| "eval_accuracy": 0.370730717794492, | |
| "eval_loss": 3.557082414627075, | |
| "eval_runtime": 82.3684, | |
| "eval_samples_per_second": 202.153, | |
| "eval_steps_per_second": 12.638, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.402305542617606, | |
| "grad_norm": 0.3769955635070801, | |
| "learning_rate": 0.00043943273150844487, | |
| "loss": 3.3039, | |
| "step": 46050 | |
| }, | |
| { | |
| "epoch": 13.416860735910573, | |
| "grad_norm": 0.37702763080596924, | |
| "learning_rate": 0.0004392580081537565, | |
| "loss": 3.301, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 13.43141592920354, | |
| "grad_norm": 0.3565526306629181, | |
| "learning_rate": 0.0004390832847990681, | |
| "loss": 3.3115, | |
| "step": 46150 | |
| }, | |
| { | |
| "epoch": 13.445971122496507, | |
| "grad_norm": 0.347365140914917, | |
| "learning_rate": 0.0004389085614443797, | |
| "loss": 3.3065, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 13.460526315789474, | |
| "grad_norm": 0.38019073009490967, | |
| "learning_rate": 0.00043873383808969125, | |
| "loss": 3.2915, | |
| "step": 46250 | |
| }, | |
| { | |
| "epoch": 13.475081509082441, | |
| "grad_norm": 0.3460626006126404, | |
| "learning_rate": 0.00043855911473500284, | |
| "loss": 3.3107, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 13.489636702375407, | |
| "grad_norm": 0.35233747959136963, | |
| "learning_rate": 0.0004383843913803145, | |
| "loss": 3.3102, | |
| "step": 46350 | |
| }, | |
| { | |
| "epoch": 13.504191895668374, | |
| "grad_norm": 0.34811362624168396, | |
| "learning_rate": 0.0004382096680256261, | |
| "loss": 3.3043, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 13.51874708896134, | |
| "grad_norm": 0.37742602825164795, | |
| "learning_rate": 0.0004380349446709376, | |
| "loss": 3.3139, | |
| "step": 46450 | |
| }, | |
| { | |
| "epoch": 13.533302282254308, | |
| "grad_norm": 0.34751424193382263, | |
| "learning_rate": 0.0004378602213162492, | |
| "loss": 3.3197, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 13.547857475547275, | |
| "grad_norm": 0.3492797613143921, | |
| "learning_rate": 0.0004376854979615608, | |
| "loss": 3.3147, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 13.562412668840242, | |
| "grad_norm": 0.34917566180229187, | |
| "learning_rate": 0.0004375107746068724, | |
| "loss": 3.3246, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 13.576967862133209, | |
| "grad_norm": 0.38131392002105713, | |
| "learning_rate": 0.000437336051252184, | |
| "loss": 3.3244, | |
| "step": 46650 | |
| }, | |
| { | |
| "epoch": 13.591523055426176, | |
| "grad_norm": 0.33767449855804443, | |
| "learning_rate": 0.0004371613278974956, | |
| "loss": 3.3171, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 13.606078248719143, | |
| "grad_norm": 0.3554423451423645, | |
| "learning_rate": 0.0004369866045428072, | |
| "loss": 3.3178, | |
| "step": 46750 | |
| }, | |
| { | |
| "epoch": 13.62063344201211, | |
| "grad_norm": 0.3402608633041382, | |
| "learning_rate": 0.0004368118811881188, | |
| "loss": 3.3162, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 13.635188635305077, | |
| "grad_norm": 0.35023102164268494, | |
| "learning_rate": 0.0004366371578334303, | |
| "loss": 3.3212, | |
| "step": 46850 | |
| }, | |
| { | |
| "epoch": 13.649743828598044, | |
| "grad_norm": 0.39269959926605225, | |
| "learning_rate": 0.000436462434478742, | |
| "loss": 3.3393, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 13.664299021891011, | |
| "grad_norm": 0.39389166235923767, | |
| "learning_rate": 0.00043628771112405357, | |
| "loss": 3.3332, | |
| "step": 46950 | |
| }, | |
| { | |
| "epoch": 13.678854215183978, | |
| "grad_norm": 0.34735947847366333, | |
| "learning_rate": 0.00043611298776936516, | |
| "loss": 3.3338, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.678854215183978, | |
| "eval_accuracy": 0.37130542631796004, | |
| "eval_loss": 3.550403356552124, | |
| "eval_runtime": 82.4411, | |
| "eval_samples_per_second": 201.975, | |
| "eval_steps_per_second": 12.627, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.693409408476944, | |
| "grad_norm": 0.34736397862434387, | |
| "learning_rate": 0.0004359382644146767, | |
| "loss": 3.3311, | |
| "step": 47050 | |
| }, | |
| { | |
| "epoch": 13.70796460176991, | |
| "grad_norm": 0.3662548363208771, | |
| "learning_rate": 0.0004357635410599883, | |
| "loss": 3.3253, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 13.722519795062878, | |
| "grad_norm": 0.36221426725387573, | |
| "learning_rate": 0.0004355888177052999, | |
| "loss": 3.3366, | |
| "step": 47150 | |
| }, | |
| { | |
| "epoch": 13.737074988355845, | |
| "grad_norm": 0.36205658316612244, | |
| "learning_rate": 0.00043541409435061154, | |
| "loss": 3.3291, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 13.751630181648812, | |
| "grad_norm": 0.3611477017402649, | |
| "learning_rate": 0.0004352393709959231, | |
| "loss": 3.3303, | |
| "step": 47250 | |
| }, | |
| { | |
| "epoch": 13.766185374941779, | |
| "grad_norm": 0.37451112270355225, | |
| "learning_rate": 0.0004350646476412347, | |
| "loss": 3.3411, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 13.780740568234746, | |
| "grad_norm": 0.33589765429496765, | |
| "learning_rate": 0.00043488992428654627, | |
| "loss": 3.3392, | |
| "step": 47350 | |
| }, | |
| { | |
| "epoch": 13.795295761527713, | |
| "grad_norm": 0.37885552644729614, | |
| "learning_rate": 0.0004347152009318578, | |
| "loss": 3.3234, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 13.80985095482068, | |
| "grad_norm": 0.3494377136230469, | |
| "learning_rate": 0.00043454047757716946, | |
| "loss": 3.3323, | |
| "step": 47450 | |
| }, | |
| { | |
| "epoch": 13.824406148113647, | |
| "grad_norm": 0.36161160469055176, | |
| "learning_rate": 0.00043436575422248105, | |
| "loss": 3.3337, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 13.838961341406614, | |
| "grad_norm": 0.3705306649208069, | |
| "learning_rate": 0.00043419103086779265, | |
| "loss": 3.3333, | |
| "step": 47550 | |
| }, | |
| { | |
| "epoch": 13.853516534699581, | |
| "grad_norm": 0.3467465043067932, | |
| "learning_rate": 0.0004340163075131042, | |
| "loss": 3.3241, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 13.868071727992549, | |
| "grad_norm": 0.4029048681259155, | |
| "learning_rate": 0.0004338415841584158, | |
| "loss": 3.3407, | |
| "step": 47650 | |
| }, | |
| { | |
| "epoch": 13.882626921285514, | |
| "grad_norm": 0.3680288791656494, | |
| "learning_rate": 0.0004336668608037274, | |
| "loss": 3.3481, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 13.89718211457848, | |
| "grad_norm": 0.34350666403770447, | |
| "learning_rate": 0.000433492137449039, | |
| "loss": 3.334, | |
| "step": 47750 | |
| }, | |
| { | |
| "epoch": 13.911737307871448, | |
| "grad_norm": 0.3675829768180847, | |
| "learning_rate": 0.00043331741409435056, | |
| "loss": 3.3277, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 13.926292501164415, | |
| "grad_norm": 0.36623910069465637, | |
| "learning_rate": 0.00043314269073966216, | |
| "loss": 3.3381, | |
| "step": 47850 | |
| }, | |
| { | |
| "epoch": 13.940847694457382, | |
| "grad_norm": 0.3612861931324005, | |
| "learning_rate": 0.00043296796738497375, | |
| "loss": 3.3273, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 13.955402887750349, | |
| "grad_norm": 0.35151249170303345, | |
| "learning_rate": 0.00043279324403028535, | |
| "loss": 3.3307, | |
| "step": 47950 | |
| }, | |
| { | |
| "epoch": 13.969958081043316, | |
| "grad_norm": 0.3364584445953369, | |
| "learning_rate": 0.0004326185206755969, | |
| "loss": 3.3351, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.969958081043316, | |
| "eval_accuracy": 0.3717133635950474, | |
| "eval_loss": 3.5407094955444336, | |
| "eval_runtime": 82.4456, | |
| "eval_samples_per_second": 201.963, | |
| "eval_steps_per_second": 12.627, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.984513274336283, | |
| "grad_norm": 0.35524892807006836, | |
| "learning_rate": 0.00043244379732090854, | |
| "loss": 3.3377, | |
| "step": 48050 | |
| }, | |
| { | |
| "epoch": 13.99906846762925, | |
| "grad_norm": 0.3291095793247223, | |
| "learning_rate": 0.00043226907396622013, | |
| "loss": 3.3519, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 14.01339077782953, | |
| "grad_norm": 0.37506425380706787, | |
| "learning_rate": 0.0004320943506115317, | |
| "loss": 3.2184, | |
| "step": 48150 | |
| }, | |
| { | |
| "epoch": 14.027945971122497, | |
| "grad_norm": 0.35535284876823425, | |
| "learning_rate": 0.00043191962725684326, | |
| "loss": 3.2396, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 14.042501164415464, | |
| "grad_norm": 0.36266106367111206, | |
| "learning_rate": 0.00043174490390215486, | |
| "loss": 3.2426, | |
| "step": 48250 | |
| }, | |
| { | |
| "epoch": 14.057056357708431, | |
| "grad_norm": 0.360951691865921, | |
| "learning_rate": 0.0004315701805474665, | |
| "loss": 3.2467, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 14.071611551001398, | |
| "grad_norm": 0.3857077360153198, | |
| "learning_rate": 0.0004313954571927781, | |
| "loss": 3.2539, | |
| "step": 48350 | |
| }, | |
| { | |
| "epoch": 14.086166744294363, | |
| "grad_norm": 0.38090914487838745, | |
| "learning_rate": 0.00043122073383808964, | |
| "loss": 3.2492, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 14.10072193758733, | |
| "grad_norm": 0.35230493545532227, | |
| "learning_rate": 0.00043104601048340124, | |
| "loss": 3.237, | |
| "step": 48450 | |
| }, | |
| { | |
| "epoch": 14.115277130880298, | |
| "grad_norm": 0.3642329275608063, | |
| "learning_rate": 0.00043087128712871283, | |
| "loss": 3.2525, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 14.129832324173265, | |
| "grad_norm": 0.3675161898136139, | |
| "learning_rate": 0.00043069656377402437, | |
| "loss": 3.2413, | |
| "step": 48550 | |
| }, | |
| { | |
| "epoch": 14.144387517466232, | |
| "grad_norm": 0.3832390606403351, | |
| "learning_rate": 0.000430521840419336, | |
| "loss": 3.2556, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 14.158942710759199, | |
| "grad_norm": 0.3640945255756378, | |
| "learning_rate": 0.0004303471170646476, | |
| "loss": 3.2672, | |
| "step": 48650 | |
| }, | |
| { | |
| "epoch": 14.173497904052166, | |
| "grad_norm": 0.37863612174987793, | |
| "learning_rate": 0.0004301723937099592, | |
| "loss": 3.264, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 14.188053097345133, | |
| "grad_norm": 0.3803202509880066, | |
| "learning_rate": 0.00042999767035527075, | |
| "loss": 3.2582, | |
| "step": 48750 | |
| }, | |
| { | |
| "epoch": 14.2026082906381, | |
| "grad_norm": 0.36837607622146606, | |
| "learning_rate": 0.00042982294700058234, | |
| "loss": 3.2634, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 14.217163483931067, | |
| "grad_norm": 0.40386468172073364, | |
| "learning_rate": 0.000429648223645894, | |
| "loss": 3.2691, | |
| "step": 48850 | |
| }, | |
| { | |
| "epoch": 14.231718677224034, | |
| "grad_norm": 0.3424507677555084, | |
| "learning_rate": 0.0004294735002912056, | |
| "loss": 3.281, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 14.246273870517001, | |
| "grad_norm": 0.3757544755935669, | |
| "learning_rate": 0.0004292987769365172, | |
| "loss": 3.2618, | |
| "step": 48950 | |
| }, | |
| { | |
| "epoch": 14.260829063809968, | |
| "grad_norm": 0.3702467083930969, | |
| "learning_rate": 0.0004291240535818287, | |
| "loss": 3.2729, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.260829063809968, | |
| "eval_accuracy": 0.37105932414042586, | |
| "eval_loss": 3.560472249984741, | |
| "eval_runtime": 82.3768, | |
| "eval_samples_per_second": 202.132, | |
| "eval_steps_per_second": 12.637, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.275384257102935, | |
| "grad_norm": 0.3377811312675476, | |
| "learning_rate": 0.0004289493302271403, | |
| "loss": 3.2804, | |
| "step": 49050 | |
| }, | |
| { | |
| "epoch": 14.2899394503959, | |
| "grad_norm": 0.3755817115306854, | |
| "learning_rate": 0.0004287746068724519, | |
| "loss": 3.2724, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 14.304494643688868, | |
| "grad_norm": 0.3477305769920349, | |
| "learning_rate": 0.00042859988351776356, | |
| "loss": 3.2875, | |
| "step": 49150 | |
| }, | |
| { | |
| "epoch": 14.319049836981835, | |
| "grad_norm": 0.3759228587150574, | |
| "learning_rate": 0.0004284251601630751, | |
| "loss": 3.295, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 14.333605030274802, | |
| "grad_norm": 0.36636483669281006, | |
| "learning_rate": 0.0004282504368083867, | |
| "loss": 3.2822, | |
| "step": 49250 | |
| }, | |
| { | |
| "epoch": 14.348160223567769, | |
| "grad_norm": 0.3604365587234497, | |
| "learning_rate": 0.0004280757134536983, | |
| "loss": 3.2849, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 14.362715416860736, | |
| "grad_norm": 0.372327595949173, | |
| "learning_rate": 0.0004279009900990098, | |
| "loss": 3.2887, | |
| "step": 49350 | |
| }, | |
| { | |
| "epoch": 14.377270610153703, | |
| "grad_norm": 0.3740226924419403, | |
| "learning_rate": 0.0004277262667443214, | |
| "loss": 3.2862, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 14.39182580344667, | |
| "grad_norm": 0.36287322640419006, | |
| "learning_rate": 0.00042755154338963307, | |
| "loss": 3.2893, | |
| "step": 49450 | |
| }, | |
| { | |
| "epoch": 14.406380996739637, | |
| "grad_norm": 0.3711443841457367, | |
| "learning_rate": 0.00042737682003494466, | |
| "loss": 3.2932, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 14.420936190032604, | |
| "grad_norm": 0.3679095208644867, | |
| "learning_rate": 0.0004272020966802562, | |
| "loss": 3.2795, | |
| "step": 49550 | |
| }, | |
| { | |
| "epoch": 14.435491383325571, | |
| "grad_norm": 0.38208362460136414, | |
| "learning_rate": 0.0004270273733255678, | |
| "loss": 3.2774, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 14.450046576618538, | |
| "grad_norm": 0.3713175356388092, | |
| "learning_rate": 0.0004268526499708794, | |
| "loss": 3.2824, | |
| "step": 49650 | |
| }, | |
| { | |
| "epoch": 14.464601769911505, | |
| "grad_norm": 0.35326892137527466, | |
| "learning_rate": 0.00042667792661619104, | |
| "loss": 3.2996, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 14.47915696320447, | |
| "grad_norm": 0.3771480321884155, | |
| "learning_rate": 0.0004265032032615026, | |
| "loss": 3.2975, | |
| "step": 49750 | |
| }, | |
| { | |
| "epoch": 14.493712156497438, | |
| "grad_norm": 0.3490285873413086, | |
| "learning_rate": 0.0004263284799068142, | |
| "loss": 3.2913, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 14.508267349790405, | |
| "grad_norm": 0.37504056096076965, | |
| "learning_rate": 0.00042615375655212577, | |
| "loss": 3.3002, | |
| "step": 49850 | |
| }, | |
| { | |
| "epoch": 14.522822543083372, | |
| "grad_norm": 0.34590810537338257, | |
| "learning_rate": 0.00042597903319743736, | |
| "loss": 3.2966, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 14.537377736376339, | |
| "grad_norm": 0.41589200496673584, | |
| "learning_rate": 0.0004258043098427489, | |
| "loss": 3.3069, | |
| "step": 49950 | |
| }, | |
| { | |
| "epoch": 14.551932929669306, | |
| "grad_norm": 0.46457570791244507, | |
| "learning_rate": 0.00042562958648806055, | |
| "loss": 3.3077, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.551932929669306, | |
| "eval_accuracy": 0.3718735533123249, | |
| "eval_loss": 3.549790859222412, | |
| "eval_runtime": 82.3343, | |
| "eval_samples_per_second": 202.236, | |
| "eval_steps_per_second": 12.644, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.566488122962273, | |
| "grad_norm": 0.34886544942855835, | |
| "learning_rate": 0.00042545486313337214, | |
| "loss": 3.306, | |
| "step": 50050 | |
| }, | |
| { | |
| "epoch": 14.58104331625524, | |
| "grad_norm": 0.34457868337631226, | |
| "learning_rate": 0.00042528013977868374, | |
| "loss": 3.3008, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 14.595598509548207, | |
| "grad_norm": 0.3627122938632965, | |
| "learning_rate": 0.0004251054164239953, | |
| "loss": 3.3121, | |
| "step": 50150 | |
| }, | |
| { | |
| "epoch": 14.610153702841174, | |
| "grad_norm": 0.3377625644207001, | |
| "learning_rate": 0.0004249306930693069, | |
| "loss": 3.3115, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 14.624708896134141, | |
| "grad_norm": 0.35513928532600403, | |
| "learning_rate": 0.0004247559697146185, | |
| "loss": 3.3071, | |
| "step": 50250 | |
| }, | |
| { | |
| "epoch": 14.639264089427108, | |
| "grad_norm": 0.36622241139411926, | |
| "learning_rate": 0.0004245812463599301, | |
| "loss": 3.3131, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 14.653819282720075, | |
| "grad_norm": 0.35609689354896545, | |
| "learning_rate": 0.00042440652300524166, | |
| "loss": 3.317, | |
| "step": 50350 | |
| }, | |
| { | |
| "epoch": 14.668374476013042, | |
| "grad_norm": 0.3505551218986511, | |
| "learning_rate": 0.00042423179965055325, | |
| "loss": 3.317, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 14.682929669306008, | |
| "grad_norm": 0.3470185697078705, | |
| "learning_rate": 0.00042405707629586484, | |
| "loss": 3.3121, | |
| "step": 50450 | |
| }, | |
| { | |
| "epoch": 14.697484862598975, | |
| "grad_norm": 0.37756529450416565, | |
| "learning_rate": 0.0004238823529411764, | |
| "loss": 3.3101, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 14.712040055891942, | |
| "grad_norm": 0.3599080741405487, | |
| "learning_rate": 0.00042370762958648803, | |
| "loss": 3.3111, | |
| "step": 50550 | |
| }, | |
| { | |
| "epoch": 14.726595249184909, | |
| "grad_norm": 0.4144226312637329, | |
| "learning_rate": 0.00042353290623179963, | |
| "loss": 3.3184, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 14.741150442477876, | |
| "grad_norm": 0.3312997817993164, | |
| "learning_rate": 0.0004233581828771112, | |
| "loss": 3.3175, | |
| "step": 50650 | |
| }, | |
| { | |
| "epoch": 14.755705635770843, | |
| "grad_norm": 0.36238470673561096, | |
| "learning_rate": 0.00042318345952242276, | |
| "loss": 3.3242, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 14.77026082906381, | |
| "grad_norm": 0.35454225540161133, | |
| "learning_rate": 0.00042300873616773436, | |
| "loss": 3.3116, | |
| "step": 50750 | |
| }, | |
| { | |
| "epoch": 14.784816022356777, | |
| "grad_norm": 0.3393413722515106, | |
| "learning_rate": 0.00042283401281304595, | |
| "loss": 3.3141, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 14.799371215649744, | |
| "grad_norm": 0.38941121101379395, | |
| "learning_rate": 0.0004226592894583576, | |
| "loss": 3.3193, | |
| "step": 50850 | |
| }, | |
| { | |
| "epoch": 14.813926408942711, | |
| "grad_norm": 0.370608389377594, | |
| "learning_rate": 0.00042248456610366914, | |
| "loss": 3.3039, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 14.828481602235678, | |
| "grad_norm": 0.3740283250808716, | |
| "learning_rate": 0.00042230984274898073, | |
| "loss": 3.314, | |
| "step": 50950 | |
| }, | |
| { | |
| "epoch": 14.843036795528645, | |
| "grad_norm": 0.3565978407859802, | |
| "learning_rate": 0.00042213511939429233, | |
| "loss": 3.3219, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.843036795528645, | |
| "eval_accuracy": 0.37199672192839744, | |
| "eval_loss": 3.5442278385162354, | |
| "eval_runtime": 82.6479, | |
| "eval_samples_per_second": 201.469, | |
| "eval_steps_per_second": 12.596, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.857591988821612, | |
| "grad_norm": 0.3865247666835785, | |
| "learning_rate": 0.0004219603960396039, | |
| "loss": 3.316, | |
| "step": 51050 | |
| }, | |
| { | |
| "epoch": 14.872147182114578, | |
| "grad_norm": 0.3700854480266571, | |
| "learning_rate": 0.0004217856726849155, | |
| "loss": 3.3261, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 14.886702375407545, | |
| "grad_norm": 0.351375937461853, | |
| "learning_rate": 0.0004216109493302271, | |
| "loss": 3.337, | |
| "step": 51150 | |
| }, | |
| { | |
| "epoch": 14.901257568700512, | |
| "grad_norm": 0.3901119828224182, | |
| "learning_rate": 0.0004214362259755387, | |
| "loss": 3.3152, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 14.915812761993479, | |
| "grad_norm": 0.3563941717147827, | |
| "learning_rate": 0.0004212615026208503, | |
| "loss": 3.3291, | |
| "step": 51250 | |
| }, | |
| { | |
| "epoch": 14.930367955286446, | |
| "grad_norm": 0.34845635294914246, | |
| "learning_rate": 0.00042108677926616184, | |
| "loss": 3.3324, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 14.944923148579413, | |
| "grad_norm": 0.3601028621196747, | |
| "learning_rate": 0.00042091205591147343, | |
| "loss": 3.3175, | |
| "step": 51350 | |
| }, | |
| { | |
| "epoch": 14.95947834187238, | |
| "grad_norm": 0.3390214741230011, | |
| "learning_rate": 0.0004207373325567851, | |
| "loss": 3.3224, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 14.974033535165347, | |
| "grad_norm": 0.3665161728858948, | |
| "learning_rate": 0.0004205626092020967, | |
| "loss": 3.3333, | |
| "step": 51450 | |
| }, | |
| { | |
| "epoch": 14.988588728458314, | |
| "grad_norm": 0.3695763051509857, | |
| "learning_rate": 0.0004203878858474082, | |
| "loss": 3.3401, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 15.002911038658594, | |
| "grad_norm": 0.35991591215133667, | |
| "learning_rate": 0.0004202131624927198, | |
| "loss": 3.3063, | |
| "step": 51550 | |
| }, | |
| { | |
| "epoch": 15.01746623195156, | |
| "grad_norm": 0.3646226227283478, | |
| "learning_rate": 0.0004200384391380314, | |
| "loss": 3.2133, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 15.032021425244528, | |
| "grad_norm": 0.37182173132896423, | |
| "learning_rate": 0.00041986371578334305, | |
| "loss": 3.2223, | |
| "step": 51650 | |
| }, | |
| { | |
| "epoch": 15.046576618537495, | |
| "grad_norm": 0.36872124671936035, | |
| "learning_rate": 0.0004196889924286546, | |
| "loss": 3.2254, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 15.06113181183046, | |
| "grad_norm": 0.38012024760246277, | |
| "learning_rate": 0.0004195142690739662, | |
| "loss": 3.2235, | |
| "step": 51750 | |
| }, | |
| { | |
| "epoch": 15.075687005123427, | |
| "grad_norm": 0.36627426743507385, | |
| "learning_rate": 0.0004193395457192778, | |
| "loss": 3.2289, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 15.090242198416394, | |
| "grad_norm": 0.37427350878715515, | |
| "learning_rate": 0.0004191648223645893, | |
| "loss": 3.2281, | |
| "step": 51850 | |
| }, | |
| { | |
| "epoch": 15.104797391709361, | |
| "grad_norm": 0.3652416467666626, | |
| "learning_rate": 0.0004189900990099009, | |
| "loss": 3.2361, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 15.119352585002328, | |
| "grad_norm": 0.38913342356681824, | |
| "learning_rate": 0.00041881537565521256, | |
| "loss": 3.2383, | |
| "step": 51950 | |
| }, | |
| { | |
| "epoch": 15.133907778295296, | |
| "grad_norm": 0.3719433844089508, | |
| "learning_rate": 0.00041864065230052416, | |
| "loss": 3.2306, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.133907778295296, | |
| "eval_accuracy": 0.371666822782104, | |
| "eval_loss": 3.5566258430480957, | |
| "eval_runtime": 82.3625, | |
| "eval_samples_per_second": 202.167, | |
| "eval_steps_per_second": 12.639, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.148462971588263, | |
| "grad_norm": 0.36681851744651794, | |
| "learning_rate": 0.00041846592894583575, | |
| "loss": 3.252, | |
| "step": 52050 | |
| }, | |
| { | |
| "epoch": 15.16301816488123, | |
| "grad_norm": 0.36448416113853455, | |
| "learning_rate": 0.0004182912055911473, | |
| "loss": 3.2452, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 15.177573358174197, | |
| "grad_norm": 0.3675693869590759, | |
| "learning_rate": 0.0004181164822364589, | |
| "loss": 3.2463, | |
| "step": 52150 | |
| }, | |
| { | |
| "epoch": 15.192128551467164, | |
| "grad_norm": 0.368488609790802, | |
| "learning_rate": 0.0004179417588817705, | |
| "loss": 3.2504, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 15.20668374476013, | |
| "grad_norm": 0.37636369466781616, | |
| "learning_rate": 0.00041776703552708213, | |
| "loss": 3.2603, | |
| "step": 52250 | |
| }, | |
| { | |
| "epoch": 15.221238938053098, | |
| "grad_norm": 0.36617031693458557, | |
| "learning_rate": 0.00041759231217239367, | |
| "loss": 3.263, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 15.235794131346065, | |
| "grad_norm": 0.3773406147956848, | |
| "learning_rate": 0.00041741758881770527, | |
| "loss": 3.2597, | |
| "step": 52350 | |
| }, | |
| { | |
| "epoch": 15.250349324639032, | |
| "grad_norm": 0.3715342581272125, | |
| "learning_rate": 0.00041724286546301686, | |
| "loss": 3.2608, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 15.264904517931997, | |
| "grad_norm": 0.3575955033302307, | |
| "learning_rate": 0.0004170681421083284, | |
| "loss": 3.2577, | |
| "step": 52450 | |
| }, | |
| { | |
| "epoch": 15.279459711224964, | |
| "grad_norm": 0.37086641788482666, | |
| "learning_rate": 0.00041689341875364005, | |
| "loss": 3.2611, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 15.294014904517931, | |
| "grad_norm": 0.3876786231994629, | |
| "learning_rate": 0.00041671869539895164, | |
| "loss": 3.266, | |
| "step": 52550 | |
| }, | |
| { | |
| "epoch": 15.308570097810899, | |
| "grad_norm": 0.37803930044174194, | |
| "learning_rate": 0.00041654397204426324, | |
| "loss": 3.2679, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 15.323125291103866, | |
| "grad_norm": 0.3776412308216095, | |
| "learning_rate": 0.0004163692486895748, | |
| "loss": 3.2761, | |
| "step": 52650 | |
| }, | |
| { | |
| "epoch": 15.337680484396833, | |
| "grad_norm": 0.363808810710907, | |
| "learning_rate": 0.00041619452533488637, | |
| "loss": 3.2719, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 15.3522356776898, | |
| "grad_norm": 0.4023805558681488, | |
| "learning_rate": 0.00041601980198019797, | |
| "loss": 3.2726, | |
| "step": 52750 | |
| }, | |
| { | |
| "epoch": 15.366790870982767, | |
| "grad_norm": 0.3668515086174011, | |
| "learning_rate": 0.0004158450786255096, | |
| "loss": 3.273, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 15.381346064275734, | |
| "grad_norm": 0.33950138092041016, | |
| "learning_rate": 0.00041567035527082115, | |
| "loss": 3.2702, | |
| "step": 52850 | |
| }, | |
| { | |
| "epoch": 15.3959012575687, | |
| "grad_norm": 0.3796953856945038, | |
| "learning_rate": 0.00041549563191613275, | |
| "loss": 3.2781, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 15.410456450861668, | |
| "grad_norm": 0.39178523421287537, | |
| "learning_rate": 0.00041532090856144434, | |
| "loss": 3.2901, | |
| "step": 52950 | |
| }, | |
| { | |
| "epoch": 15.425011644154635, | |
| "grad_norm": 0.3470476269721985, | |
| "learning_rate": 0.00041514618520675594, | |
| "loss": 3.2878, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.425011644154635, | |
| "eval_accuracy": 0.37158713926903425, | |
| "eval_loss": 3.551004409790039, | |
| "eval_runtime": 82.3253, | |
| "eval_samples_per_second": 202.259, | |
| "eval_steps_per_second": 12.645, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.439566837447602, | |
| "grad_norm": 0.3517593443393707, | |
| "learning_rate": 0.00041497146185206753, | |
| "loss": 3.274, | |
| "step": 53050 | |
| }, | |
| { | |
| "epoch": 15.454122030740567, | |
| "grad_norm": 0.3858201503753662, | |
| "learning_rate": 0.0004147967384973791, | |
| "loss": 3.2776, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 15.468677224033534, | |
| "grad_norm": 0.37728390097618103, | |
| "learning_rate": 0.0004146220151426907, | |
| "loss": 3.2798, | |
| "step": 53150 | |
| }, | |
| { | |
| "epoch": 15.483232417326501, | |
| "grad_norm": 0.3583352863788605, | |
| "learning_rate": 0.0004144472917880023, | |
| "loss": 3.2805, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 15.497787610619469, | |
| "grad_norm": 0.34163346886634827, | |
| "learning_rate": 0.00041427256843331385, | |
| "loss": 3.2918, | |
| "step": 53250 | |
| }, | |
| { | |
| "epoch": 15.512342803912436, | |
| "grad_norm": 0.3739677667617798, | |
| "learning_rate": 0.00041409784507862545, | |
| "loss": 3.2914, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 15.526897997205403, | |
| "grad_norm": 0.35051360726356506, | |
| "learning_rate": 0.0004139231217239371, | |
| "loss": 3.2689, | |
| "step": 53350 | |
| }, | |
| { | |
| "epoch": 15.54145319049837, | |
| "grad_norm": 0.3869415819644928, | |
| "learning_rate": 0.0004137483983692487, | |
| "loss": 3.2911, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 15.556008383791337, | |
| "grad_norm": 0.36137157678604126, | |
| "learning_rate": 0.00041357367501456023, | |
| "loss": 3.2854, | |
| "step": 53450 | |
| }, | |
| { | |
| "epoch": 15.570563577084304, | |
| "grad_norm": 0.39776507019996643, | |
| "learning_rate": 0.0004133989516598718, | |
| "loss": 3.2824, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 15.585118770377271, | |
| "grad_norm": 0.38777944445610046, | |
| "learning_rate": 0.0004132242283051834, | |
| "loss": 3.2965, | |
| "step": 53550 | |
| }, | |
| { | |
| "epoch": 15.599673963670238, | |
| "grad_norm": 0.3499772250652313, | |
| "learning_rate": 0.00041304950495049496, | |
| "loss": 3.3, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 15.614229156963205, | |
| "grad_norm": 0.36362919211387634, | |
| "learning_rate": 0.0004128747815958066, | |
| "loss": 3.3019, | |
| "step": 53650 | |
| }, | |
| { | |
| "epoch": 15.628784350256172, | |
| "grad_norm": 0.3737014830112457, | |
| "learning_rate": 0.0004127000582411182, | |
| "loss": 3.2973, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 15.64333954354914, | |
| "grad_norm": 0.3679458200931549, | |
| "learning_rate": 0.0004125253348864298, | |
| "loss": 3.2951, | |
| "step": 53750 | |
| }, | |
| { | |
| "epoch": 15.657894736842106, | |
| "grad_norm": 0.38124388456344604, | |
| "learning_rate": 0.00041235061153174134, | |
| "loss": 3.299, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 15.672449930135071, | |
| "grad_norm": 0.37258169054985046, | |
| "learning_rate": 0.00041217588817705293, | |
| "loss": 3.2852, | |
| "step": 53850 | |
| }, | |
| { | |
| "epoch": 15.687005123428039, | |
| "grad_norm": 0.37887269258499146, | |
| "learning_rate": 0.0004120011648223646, | |
| "loss": 3.306, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 15.701560316721006, | |
| "grad_norm": 0.3725917637348175, | |
| "learning_rate": 0.0004118264414676762, | |
| "loss": 3.2876, | |
| "step": 53950 | |
| }, | |
| { | |
| "epoch": 15.716115510013973, | |
| "grad_norm": 0.3422054052352905, | |
| "learning_rate": 0.0004116517181129877, | |
| "loss": 3.2904, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.716115510013973, | |
| "eval_accuracy": 0.37227208840497933, | |
| "eval_loss": 3.544064521789551, | |
| "eval_runtime": 82.3559, | |
| "eval_samples_per_second": 202.184, | |
| "eval_steps_per_second": 12.64, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.73067070330694, | |
| "grad_norm": 0.3626137971878052, | |
| "learning_rate": 0.0004114769947582993, | |
| "loss": 3.2952, | |
| "step": 54050 | |
| }, | |
| { | |
| "epoch": 15.745225896599907, | |
| "grad_norm": 0.3450908958911896, | |
| "learning_rate": 0.0004113022714036109, | |
| "loss": 3.3037, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 15.759781089892874, | |
| "grad_norm": 0.3478912115097046, | |
| "learning_rate": 0.0004111275480489225, | |
| "loss": 3.2993, | |
| "step": 54150 | |
| }, | |
| { | |
| "epoch": 15.774336283185841, | |
| "grad_norm": 0.344109445810318, | |
| "learning_rate": 0.0004109528246942341, | |
| "loss": 3.3042, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 15.788891476478808, | |
| "grad_norm": 0.346696674823761, | |
| "learning_rate": 0.0004107781013395457, | |
| "loss": 3.3031, | |
| "step": 54250 | |
| }, | |
| { | |
| "epoch": 15.803446669771775, | |
| "grad_norm": 0.38119471073150635, | |
| "learning_rate": 0.0004106033779848573, | |
| "loss": 3.3028, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 15.818001863064742, | |
| "grad_norm": 0.3529111444950104, | |
| "learning_rate": 0.0004104286546301689, | |
| "loss": 3.3105, | |
| "step": 54350 | |
| }, | |
| { | |
| "epoch": 15.83255705635771, | |
| "grad_norm": 0.3672657608985901, | |
| "learning_rate": 0.0004102539312754804, | |
| "loss": 3.3003, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 15.847112249650674, | |
| "grad_norm": 0.35365378856658936, | |
| "learning_rate": 0.00041007920792079206, | |
| "loss": 3.3082, | |
| "step": 54450 | |
| }, | |
| { | |
| "epoch": 15.861667442943642, | |
| "grad_norm": 0.34256818890571594, | |
| "learning_rate": 0.00040990448456610366, | |
| "loss": 3.3145, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 15.876222636236609, | |
| "grad_norm": 0.3619665801525116, | |
| "learning_rate": 0.00040972976121141525, | |
| "loss": 3.3057, | |
| "step": 54550 | |
| }, | |
| { | |
| "epoch": 15.890777829529576, | |
| "grad_norm": 0.3698698878288269, | |
| "learning_rate": 0.0004095550378567268, | |
| "loss": 3.3074, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 15.905333022822543, | |
| "grad_norm": 0.3723785877227783, | |
| "learning_rate": 0.0004093803145020384, | |
| "loss": 3.3138, | |
| "step": 54650 | |
| }, | |
| { | |
| "epoch": 15.91988821611551, | |
| "grad_norm": 0.38027748465538025, | |
| "learning_rate": 0.00040920559114735, | |
| "loss": 3.3087, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 15.934443409408477, | |
| "grad_norm": 0.3341813087463379, | |
| "learning_rate": 0.00040903086779266163, | |
| "loss": 3.3194, | |
| "step": 54750 | |
| }, | |
| { | |
| "epoch": 15.948998602701444, | |
| "grad_norm": 0.3505150079727173, | |
| "learning_rate": 0.00040885614443797317, | |
| "loss": 3.3169, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 15.963553795994411, | |
| "grad_norm": 0.3617570400238037, | |
| "learning_rate": 0.00040868142108328476, | |
| "loss": 3.3138, | |
| "step": 54850 | |
| }, | |
| { | |
| "epoch": 15.978108989287378, | |
| "grad_norm": 0.4044189155101776, | |
| "learning_rate": 0.00040850669772859636, | |
| "loss": 3.3055, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 15.992664182580345, | |
| "grad_norm": 0.348711222410202, | |
| "learning_rate": 0.0004083319743739079, | |
| "loss": 3.3095, | |
| "step": 54950 | |
| }, | |
| { | |
| "epoch": 16.006986492780623, | |
| "grad_norm": 0.364349365234375, | |
| "learning_rate": 0.0004081572510192195, | |
| "loss": 3.2623, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.006986492780623, | |
| "eval_accuracy": 0.37194406969557253, | |
| "eval_loss": 3.5528974533081055, | |
| "eval_runtime": 82.327, | |
| "eval_samples_per_second": 202.254, | |
| "eval_steps_per_second": 12.645, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.02154168607359, | |
| "grad_norm": 0.35028260946273804, | |
| "learning_rate": 0.00040798252766453114, | |
| "loss": 3.206, | |
| "step": 55050 | |
| }, | |
| { | |
| "epoch": 16.036096879366557, | |
| "grad_norm": 0.36651697754859924, | |
| "learning_rate": 0.00040780780430984273, | |
| "loss": 3.2009, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 16.050652072659524, | |
| "grad_norm": 0.35472917556762695, | |
| "learning_rate": 0.0004076330809551543, | |
| "loss": 3.2042, | |
| "step": 55150 | |
| }, | |
| { | |
| "epoch": 16.06520726595249, | |
| "grad_norm": 0.3583957552909851, | |
| "learning_rate": 0.00040745835760046587, | |
| "loss": 3.219, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 16.079762459245458, | |
| "grad_norm": 0.3749157786369324, | |
| "learning_rate": 0.00040728363424577746, | |
| "loss": 3.2086, | |
| "step": 55250 | |
| }, | |
| { | |
| "epoch": 16.094317652538425, | |
| "grad_norm": 0.3825204372406006, | |
| "learning_rate": 0.0004071089108910891, | |
| "loss": 3.2298, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 16.108872845831392, | |
| "grad_norm": 0.3701384961605072, | |
| "learning_rate": 0.0004069341875364007, | |
| "loss": 3.2346, | |
| "step": 55350 | |
| }, | |
| { | |
| "epoch": 16.12342803912436, | |
| "grad_norm": 0.3834715187549591, | |
| "learning_rate": 0.00040675946418171225, | |
| "loss": 3.2216, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 16.137983232417326, | |
| "grad_norm": 0.3612671196460724, | |
| "learning_rate": 0.00040658474082702384, | |
| "loss": 3.2248, | |
| "step": 55450 | |
| }, | |
| { | |
| "epoch": 16.152538425710294, | |
| "grad_norm": 0.36859214305877686, | |
| "learning_rate": 0.00040641001747233543, | |
| "loss": 3.2271, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 16.16709361900326, | |
| "grad_norm": 0.3895788788795471, | |
| "learning_rate": 0.000406235294117647, | |
| "loss": 3.2212, | |
| "step": 55550 | |
| }, | |
| { | |
| "epoch": 16.181648812296228, | |
| "grad_norm": 0.3740449845790863, | |
| "learning_rate": 0.0004060605707629586, | |
| "loss": 3.2355, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 16.196204005589195, | |
| "grad_norm": 0.38083046674728394, | |
| "learning_rate": 0.0004058858474082702, | |
| "loss": 3.2451, | |
| "step": 55650 | |
| }, | |
| { | |
| "epoch": 16.21075919888216, | |
| "grad_norm": 0.3648262321949005, | |
| "learning_rate": 0.0004057111240535818, | |
| "loss": 3.2479, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 16.22531439217513, | |
| "grad_norm": 0.36412709951400757, | |
| "learning_rate": 0.00040553640069889335, | |
| "loss": 3.2404, | |
| "step": 55750 | |
| }, | |
| { | |
| "epoch": 16.239869585468096, | |
| "grad_norm": 0.36129316687583923, | |
| "learning_rate": 0.00040536167734420495, | |
| "loss": 3.258, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 16.254424778761063, | |
| "grad_norm": 0.36568403244018555, | |
| "learning_rate": 0.0004051869539895166, | |
| "loss": 3.2506, | |
| "step": 55850 | |
| }, | |
| { | |
| "epoch": 16.26897997205403, | |
| "grad_norm": 0.35923558473587036, | |
| "learning_rate": 0.0004050122306348282, | |
| "loss": 3.2431, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 16.283535165346997, | |
| "grad_norm": 0.36943888664245605, | |
| "learning_rate": 0.00040483750728013973, | |
| "loss": 3.2505, | |
| "step": 55950 | |
| }, | |
| { | |
| "epoch": 16.298090358639964, | |
| "grad_norm": 0.38828739523887634, | |
| "learning_rate": 0.0004046627839254513, | |
| "loss": 3.2507, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.298090358639964, | |
| "eval_accuracy": 0.3722270754470063, | |
| "eval_loss": 3.549487829208374, | |
| "eval_runtime": 82.4249, | |
| "eval_samples_per_second": 202.014, | |
| "eval_steps_per_second": 12.63, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.31264555193293, | |
| "grad_norm": 0.3798963129520416, | |
| "learning_rate": 0.0004044880605707629, | |
| "loss": 3.2594, | |
| "step": 56050 | |
| }, | |
| { | |
| "epoch": 16.3272007452259, | |
| "grad_norm": 0.35548266768455505, | |
| "learning_rate": 0.00040431333721607446, | |
| "loss": 3.2628, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 16.341755938518865, | |
| "grad_norm": 0.3860792815685272, | |
| "learning_rate": 0.0004041386138613861, | |
| "loss": 3.2574, | |
| "step": 56150 | |
| }, | |
| { | |
| "epoch": 16.35631113181183, | |
| "grad_norm": 0.3871467709541321, | |
| "learning_rate": 0.0004039638905066977, | |
| "loss": 3.2617, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 16.370866325104796, | |
| "grad_norm": 0.36527544260025024, | |
| "learning_rate": 0.0004037891671520093, | |
| "loss": 3.2521, | |
| "step": 56250 | |
| }, | |
| { | |
| "epoch": 16.385421518397763, | |
| "grad_norm": 0.3756539523601532, | |
| "learning_rate": 0.0004036144437973209, | |
| "loss": 3.2498, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 16.39997671169073, | |
| "grad_norm": 0.35087254643440247, | |
| "learning_rate": 0.00040343972044263243, | |
| "loss": 3.2673, | |
| "step": 56350 | |
| }, | |
| { | |
| "epoch": 16.414531904983697, | |
| "grad_norm": 0.36675381660461426, | |
| "learning_rate": 0.0004032649970879441, | |
| "loss": 3.2632, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 16.429087098276664, | |
| "grad_norm": 0.3826260566711426, | |
| "learning_rate": 0.00040309027373325567, | |
| "loss": 3.2648, | |
| "step": 56450 | |
| }, | |
| { | |
| "epoch": 16.44364229156963, | |
| "grad_norm": 0.3628179430961609, | |
| "learning_rate": 0.00040291555037856727, | |
| "loss": 3.2611, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 16.4581974848626, | |
| "grad_norm": 0.3654247224330902, | |
| "learning_rate": 0.0004027408270238788, | |
| "loss": 3.2662, | |
| "step": 56550 | |
| }, | |
| { | |
| "epoch": 16.472752678155565, | |
| "grad_norm": 0.39291977882385254, | |
| "learning_rate": 0.0004025661036691904, | |
| "loss": 3.2832, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 16.487307871448532, | |
| "grad_norm": 0.4162214994430542, | |
| "learning_rate": 0.000402391380314502, | |
| "loss": 3.2759, | |
| "step": 56650 | |
| }, | |
| { | |
| "epoch": 16.5018630647415, | |
| "grad_norm": 0.36966341733932495, | |
| "learning_rate": 0.00040221665695981364, | |
| "loss": 3.2742, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 16.516418258034467, | |
| "grad_norm": 0.384409636259079, | |
| "learning_rate": 0.0004020419336051252, | |
| "loss": 3.2726, | |
| "step": 56750 | |
| }, | |
| { | |
| "epoch": 16.530973451327434, | |
| "grad_norm": 0.36694660782814026, | |
| "learning_rate": 0.0004018672102504368, | |
| "loss": 3.2775, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 16.5455286446204, | |
| "grad_norm": 0.41723883152008057, | |
| "learning_rate": 0.00040169248689574837, | |
| "loss": 3.287, | |
| "step": 56850 | |
| }, | |
| { | |
| "epoch": 16.560083837913368, | |
| "grad_norm": 0.3759129047393799, | |
| "learning_rate": 0.0004015177635410599, | |
| "loss": 3.2814, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 16.574639031206335, | |
| "grad_norm": 0.372667133808136, | |
| "learning_rate": 0.0004013430401863715, | |
| "loss": 3.2796, | |
| "step": 56950 | |
| }, | |
| { | |
| "epoch": 16.589194224499302, | |
| "grad_norm": 0.37249433994293213, | |
| "learning_rate": 0.00040116831683168315, | |
| "loss": 3.2749, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.589194224499302, | |
| "eval_accuracy": 0.3728302255783842, | |
| "eval_loss": 3.5436081886291504, | |
| "eval_runtime": 82.3525, | |
| "eval_samples_per_second": 202.192, | |
| "eval_steps_per_second": 12.641, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.60374941779227, | |
| "grad_norm": 0.3639916479587555, | |
| "learning_rate": 0.00040099359347699475, | |
| "loss": 3.2701, | |
| "step": 57050 | |
| }, | |
| { | |
| "epoch": 16.618304611085236, | |
| "grad_norm": 0.36057543754577637, | |
| "learning_rate": 0.0004008188701223063, | |
| "loss": 3.2677, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 16.632859804378203, | |
| "grad_norm": 0.3790561258792877, | |
| "learning_rate": 0.0004006441467676179, | |
| "loss": 3.2821, | |
| "step": 57150 | |
| }, | |
| { | |
| "epoch": 16.64741499767117, | |
| "grad_norm": 0.3740590810775757, | |
| "learning_rate": 0.0004004694234129295, | |
| "loss": 3.2901, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 16.661970190964137, | |
| "grad_norm": 0.35806769132614136, | |
| "learning_rate": 0.0004002947000582411, | |
| "loss": 3.2889, | |
| "step": 57250 | |
| }, | |
| { | |
| "epoch": 16.676525384257104, | |
| "grad_norm": 0.38618093729019165, | |
| "learning_rate": 0.00040011997670355267, | |
| "loss": 3.2796, | |
| "step": 57300 | |
| }, | |
| { | |
| "epoch": 16.69108057755007, | |
| "grad_norm": 0.38576868176460266, | |
| "learning_rate": 0.00039994525334886426, | |
| "loss": 3.2891, | |
| "step": 57350 | |
| }, | |
| { | |
| "epoch": 16.70563577084304, | |
| "grad_norm": 0.4014085531234741, | |
| "learning_rate": 0.00039977052999417585, | |
| "loss": 3.2956, | |
| "step": 57400 | |
| }, | |
| { | |
| "epoch": 16.720190964136005, | |
| "grad_norm": 0.37516024708747864, | |
| "learning_rate": 0.00039959580663948745, | |
| "loss": 3.2838, | |
| "step": 57450 | |
| }, | |
| { | |
| "epoch": 16.734746157428972, | |
| "grad_norm": 0.3815675377845764, | |
| "learning_rate": 0.000399421083284799, | |
| "loss": 3.2827, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 16.749301350721936, | |
| "grad_norm": 0.43055450916290283, | |
| "learning_rate": 0.00039924635993011064, | |
| "loss": 3.2914, | |
| "step": 57550 | |
| }, | |
| { | |
| "epoch": 16.763856544014903, | |
| "grad_norm": 0.385409414768219, | |
| "learning_rate": 0.00039907163657542223, | |
| "loss": 3.2985, | |
| "step": 57600 | |
| }, | |
| { | |
| "epoch": 16.77841173730787, | |
| "grad_norm": 0.3660253882408142, | |
| "learning_rate": 0.0003988969132207338, | |
| "loss": 3.2867, | |
| "step": 57650 | |
| }, | |
| { | |
| "epoch": 16.792966930600837, | |
| "grad_norm": 0.3687756061553955, | |
| "learning_rate": 0.00039872218986604537, | |
| "loss": 3.2857, | |
| "step": 57700 | |
| }, | |
| { | |
| "epoch": 16.807522123893804, | |
| "grad_norm": 0.3709680140018463, | |
| "learning_rate": 0.00039854746651135696, | |
| "loss": 3.2834, | |
| "step": 57750 | |
| }, | |
| { | |
| "epoch": 16.82207731718677, | |
| "grad_norm": 0.3358961343765259, | |
| "learning_rate": 0.0003983727431566686, | |
| "loss": 3.2891, | |
| "step": 57800 | |
| }, | |
| { | |
| "epoch": 16.83663251047974, | |
| "grad_norm": 0.37720680236816406, | |
| "learning_rate": 0.0003981980198019802, | |
| "loss": 3.292, | |
| "step": 57850 | |
| }, | |
| { | |
| "epoch": 16.851187703772705, | |
| "grad_norm": 0.38169723749160767, | |
| "learning_rate": 0.00039802329644729174, | |
| "loss": 3.3008, | |
| "step": 57900 | |
| }, | |
| { | |
| "epoch": 16.865742897065672, | |
| "grad_norm": 0.4024460017681122, | |
| "learning_rate": 0.00039784857309260334, | |
| "loss": 3.2938, | |
| "step": 57950 | |
| }, | |
| { | |
| "epoch": 16.88029809035864, | |
| "grad_norm": 0.3864589333534241, | |
| "learning_rate": 0.00039767384973791493, | |
| "loss": 3.3053, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.88029809035864, | |
| "eval_accuracy": 0.37272550874926147, | |
| "eval_loss": 3.5374245643615723, | |
| "eval_runtime": 82.3463, | |
| "eval_samples_per_second": 202.207, | |
| "eval_steps_per_second": 12.642, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.894853283651607, | |
| "grad_norm": 0.36863377690315247, | |
| "learning_rate": 0.00039749912638322647, | |
| "loss": 3.3084, | |
| "step": 58050 | |
| }, | |
| { | |
| "epoch": 16.909408476944574, | |
| "grad_norm": 0.35396283864974976, | |
| "learning_rate": 0.0003973244030285381, | |
| "loss": 3.2981, | |
| "step": 58100 | |
| }, | |
| { | |
| "epoch": 16.92396367023754, | |
| "grad_norm": 0.42627930641174316, | |
| "learning_rate": 0.0003971496796738497, | |
| "loss": 3.2983, | |
| "step": 58150 | |
| }, | |
| { | |
| "epoch": 16.938518863530508, | |
| "grad_norm": 0.38064268231391907, | |
| "learning_rate": 0.0003969749563191613, | |
| "loss": 3.3052, | |
| "step": 58200 | |
| }, | |
| { | |
| "epoch": 16.953074056823475, | |
| "grad_norm": 0.4481252133846283, | |
| "learning_rate": 0.00039680023296447285, | |
| "loss": 3.2938, | |
| "step": 58250 | |
| }, | |
| { | |
| "epoch": 16.967629250116442, | |
| "grad_norm": 0.4366079866886139, | |
| "learning_rate": 0.00039662550960978444, | |
| "loss": 3.2942, | |
| "step": 58300 | |
| }, | |
| { | |
| "epoch": 16.98218444340941, | |
| "grad_norm": 0.348246693611145, | |
| "learning_rate": 0.00039645078625509604, | |
| "loss": 3.2985, | |
| "step": 58350 | |
| }, | |
| { | |
| "epoch": 16.996739636702376, | |
| "grad_norm": 0.37742879986763, | |
| "learning_rate": 0.0003962760629004077, | |
| "loss": 3.2906, | |
| "step": 58400 | |
| }, | |
| { | |
| "epoch": 17.011061946902654, | |
| "grad_norm": 0.4014660716056824, | |
| "learning_rate": 0.0003961013395457193, | |
| "loss": 3.2176, | |
| "step": 58450 | |
| }, | |
| { | |
| "epoch": 17.02561714019562, | |
| "grad_norm": 0.3702877461910248, | |
| "learning_rate": 0.0003959266161910308, | |
| "loss": 3.1857, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 17.040172333488588, | |
| "grad_norm": 0.3448271155357361, | |
| "learning_rate": 0.0003957518928363424, | |
| "loss": 3.1976, | |
| "step": 58550 | |
| }, | |
| { | |
| "epoch": 17.054727526781555, | |
| "grad_norm": 0.41677457094192505, | |
| "learning_rate": 0.000395577169481654, | |
| "loss": 3.1962, | |
| "step": 58600 | |
| }, | |
| { | |
| "epoch": 17.069282720074522, | |
| "grad_norm": 0.3849838972091675, | |
| "learning_rate": 0.00039540244612696566, | |
| "loss": 3.2027, | |
| "step": 58650 | |
| }, | |
| { | |
| "epoch": 17.08383791336749, | |
| "grad_norm": 0.3862682580947876, | |
| "learning_rate": 0.0003952277227722772, | |
| "loss": 3.1925, | |
| "step": 58700 | |
| }, | |
| { | |
| "epoch": 17.098393106660456, | |
| "grad_norm": 0.37610191106796265, | |
| "learning_rate": 0.0003950529994175888, | |
| "loss": 3.212, | |
| "step": 58750 | |
| }, | |
| { | |
| "epoch": 17.112948299953423, | |
| "grad_norm": 0.3856978714466095, | |
| "learning_rate": 0.0003948782760629004, | |
| "loss": 3.2105, | |
| "step": 58800 | |
| }, | |
| { | |
| "epoch": 17.12750349324639, | |
| "grad_norm": 0.3871050477027893, | |
| "learning_rate": 0.0003947035527082119, | |
| "loss": 3.2079, | |
| "step": 58850 | |
| }, | |
| { | |
| "epoch": 17.142058686539357, | |
| "grad_norm": 0.38416802883148193, | |
| "learning_rate": 0.0003945288293535235, | |
| "loss": 3.2151, | |
| "step": 58900 | |
| }, | |
| { | |
| "epoch": 17.156613879832324, | |
| "grad_norm": 0.37692973017692566, | |
| "learning_rate": 0.00039435410599883517, | |
| "loss": 3.2273, | |
| "step": 58950 | |
| }, | |
| { | |
| "epoch": 17.17116907312529, | |
| "grad_norm": 0.38321807980537415, | |
| "learning_rate": 0.00039417938264414676, | |
| "loss": 3.2227, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.17116907312529, | |
| "eval_accuracy": 0.37236834326811236, | |
| "eval_loss": 3.5503973960876465, | |
| "eval_runtime": 82.3992, | |
| "eval_samples_per_second": 202.077, | |
| "eval_steps_per_second": 12.634, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.18572426641826, | |
| "grad_norm": 0.3697952330112457, | |
| "learning_rate": 0.0003940046592894583, | |
| "loss": 3.235, | |
| "step": 59050 | |
| }, | |
| { | |
| "epoch": 17.200279459711226, | |
| "grad_norm": 0.39189496636390686, | |
| "learning_rate": 0.0003938299359347699, | |
| "loss": 3.2408, | |
| "step": 59100 | |
| }, | |
| { | |
| "epoch": 17.214834653004193, | |
| "grad_norm": 0.36710870265960693, | |
| "learning_rate": 0.0003936552125800815, | |
| "loss": 3.223, | |
| "step": 59150 | |
| }, | |
| { | |
| "epoch": 17.22938984629716, | |
| "grad_norm": 0.38866865634918213, | |
| "learning_rate": 0.00039348048922539314, | |
| "loss": 3.2332, | |
| "step": 59200 | |
| }, | |
| { | |
| "epoch": 17.243945039590127, | |
| "grad_norm": 0.4000551998615265, | |
| "learning_rate": 0.0003933057658707047, | |
| "loss": 3.2212, | |
| "step": 59250 | |
| }, | |
| { | |
| "epoch": 17.258500232883094, | |
| "grad_norm": 0.39317259192466736, | |
| "learning_rate": 0.0003931310425160163, | |
| "loss": 3.234, | |
| "step": 59300 | |
| }, | |
| { | |
| "epoch": 17.27305542617606, | |
| "grad_norm": 0.39115262031555176, | |
| "learning_rate": 0.00039295631916132787, | |
| "loss": 3.2306, | |
| "step": 59350 | |
| }, | |
| { | |
| "epoch": 17.287610619469028, | |
| "grad_norm": 0.3673313558101654, | |
| "learning_rate": 0.00039278159580663946, | |
| "loss": 3.2453, | |
| "step": 59400 | |
| }, | |
| { | |
| "epoch": 17.302165812761995, | |
| "grad_norm": 0.36992019414901733, | |
| "learning_rate": 0.000392606872451951, | |
| "loss": 3.2576, | |
| "step": 59450 | |
| }, | |
| { | |
| "epoch": 17.316721006054962, | |
| "grad_norm": 0.39418211579322815, | |
| "learning_rate": 0.00039243214909726265, | |
| "loss": 3.2397, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 17.331276199347926, | |
| "grad_norm": 0.370551735162735, | |
| "learning_rate": 0.00039225742574257425, | |
| "loss": 3.2404, | |
| "step": 59550 | |
| }, | |
| { | |
| "epoch": 17.345831392640893, | |
| "grad_norm": 0.38773655891418457, | |
| "learning_rate": 0.00039208270238788584, | |
| "loss": 3.2532, | |
| "step": 59600 | |
| }, | |
| { | |
| "epoch": 17.36038658593386, | |
| "grad_norm": 0.3710317313671112, | |
| "learning_rate": 0.0003919079790331974, | |
| "loss": 3.2527, | |
| "step": 59650 | |
| }, | |
| { | |
| "epoch": 17.374941779226827, | |
| "grad_norm": 0.377973347902298, | |
| "learning_rate": 0.000391733255678509, | |
| "loss": 3.2466, | |
| "step": 59700 | |
| }, | |
| { | |
| "epoch": 17.389496972519794, | |
| "grad_norm": 0.3749079704284668, | |
| "learning_rate": 0.00039155853232382057, | |
| "loss": 3.2464, | |
| "step": 59750 | |
| }, | |
| { | |
| "epoch": 17.40405216581276, | |
| "grad_norm": 0.38407987356185913, | |
| "learning_rate": 0.0003913838089691322, | |
| "loss": 3.2486, | |
| "step": 59800 | |
| }, | |
| { | |
| "epoch": 17.418607359105728, | |
| "grad_norm": 0.38260164856910706, | |
| "learning_rate": 0.00039120908561444376, | |
| "loss": 3.2568, | |
| "step": 59850 | |
| }, | |
| { | |
| "epoch": 17.433162552398695, | |
| "grad_norm": 0.368882954120636, | |
| "learning_rate": 0.00039103436225975535, | |
| "loss": 3.2482, | |
| "step": 59900 | |
| }, | |
| { | |
| "epoch": 17.447717745691662, | |
| "grad_norm": 0.3901490867137909, | |
| "learning_rate": 0.00039085963890506695, | |
| "loss": 3.2435, | |
| "step": 59950 | |
| }, | |
| { | |
| "epoch": 17.46227293898463, | |
| "grad_norm": 0.3756331503391266, | |
| "learning_rate": 0.0003906849155503785, | |
| "loss": 3.2698, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.46227293898463, | |
| "eval_accuracy": 0.37248963144729824, | |
| "eval_loss": 3.547746181488037, | |
| "eval_runtime": 82.3256, | |
| "eval_samples_per_second": 202.258, | |
| "eval_steps_per_second": 12.645, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.476828132277596, | |
| "grad_norm": 0.36771896481513977, | |
| "learning_rate": 0.00039051019219569014, | |
| "loss": 3.2501, | |
| "step": 60050 | |
| }, | |
| { | |
| "epoch": 17.491383325570563, | |
| "grad_norm": 0.42868348956108093, | |
| "learning_rate": 0.00039033546884100173, | |
| "loss": 3.2559, | |
| "step": 60100 | |
| }, | |
| { | |
| "epoch": 17.50593851886353, | |
| "grad_norm": 0.39367952942848206, | |
| "learning_rate": 0.0003901607454863133, | |
| "loss": 3.2568, | |
| "step": 60150 | |
| }, | |
| { | |
| "epoch": 17.520493712156497, | |
| "grad_norm": 0.38117673993110657, | |
| "learning_rate": 0.00038998602213162486, | |
| "loss": 3.262, | |
| "step": 60200 | |
| }, | |
| { | |
| "epoch": 17.535048905449464, | |
| "grad_norm": 0.38053274154663086, | |
| "learning_rate": 0.00038981129877693646, | |
| "loss": 3.2757, | |
| "step": 60250 | |
| }, | |
| { | |
| "epoch": 17.54960409874243, | |
| "grad_norm": 0.39092931151390076, | |
| "learning_rate": 0.00038963657542224805, | |
| "loss": 3.2621, | |
| "step": 60300 | |
| }, | |
| { | |
| "epoch": 17.5641592920354, | |
| "grad_norm": 0.3721652626991272, | |
| "learning_rate": 0.0003894618520675597, | |
| "loss": 3.282, | |
| "step": 60350 | |
| }, | |
| { | |
| "epoch": 17.578714485328366, | |
| "grad_norm": 0.36930033564567566, | |
| "learning_rate": 0.00038928712871287124, | |
| "loss": 3.2656, | |
| "step": 60400 | |
| }, | |
| { | |
| "epoch": 17.593269678621333, | |
| "grad_norm": 0.3660658597946167, | |
| "learning_rate": 0.00038911240535818284, | |
| "loss": 3.2582, | |
| "step": 60450 | |
| }, | |
| { | |
| "epoch": 17.6078248719143, | |
| "grad_norm": 0.3680090606212616, | |
| "learning_rate": 0.00038893768200349443, | |
| "loss": 3.2665, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 17.622380065207267, | |
| "grad_norm": 0.4176238179206848, | |
| "learning_rate": 0.000388762958648806, | |
| "loss": 3.2647, | |
| "step": 60550 | |
| }, | |
| { | |
| "epoch": 17.636935258500234, | |
| "grad_norm": 0.37925779819488525, | |
| "learning_rate": 0.0003885882352941176, | |
| "loss": 3.2576, | |
| "step": 60600 | |
| }, | |
| { | |
| "epoch": 17.6514904517932, | |
| "grad_norm": 0.38179367780685425, | |
| "learning_rate": 0.0003884135119394292, | |
| "loss": 3.2713, | |
| "step": 60650 | |
| }, | |
| { | |
| "epoch": 17.666045645086168, | |
| "grad_norm": 0.382871150970459, | |
| "learning_rate": 0.0003882387885847408, | |
| "loss": 3.266, | |
| "step": 60700 | |
| }, | |
| { | |
| "epoch": 17.680600838379135, | |
| "grad_norm": 0.37724173069000244, | |
| "learning_rate": 0.0003880640652300524, | |
| "loss": 3.2643, | |
| "step": 60750 | |
| }, | |
| { | |
| "epoch": 17.695156031672102, | |
| "grad_norm": 0.39893394708633423, | |
| "learning_rate": 0.00038788934187536394, | |
| "loss": 3.2886, | |
| "step": 60800 | |
| }, | |
| { | |
| "epoch": 17.70971122496507, | |
| "grad_norm": 0.3657815158367157, | |
| "learning_rate": 0.00038771461852067554, | |
| "loss": 3.2824, | |
| "step": 60850 | |
| }, | |
| { | |
| "epoch": 17.724266418258033, | |
| "grad_norm": 0.3831231892108917, | |
| "learning_rate": 0.0003875398951659872, | |
| "loss": 3.276, | |
| "step": 60900 | |
| }, | |
| { | |
| "epoch": 17.738821611551, | |
| "grad_norm": 0.372113436460495, | |
| "learning_rate": 0.0003873651718112988, | |
| "loss": 3.2687, | |
| "step": 60950 | |
| }, | |
| { | |
| "epoch": 17.753376804843967, | |
| "grad_norm": 0.37759917974472046, | |
| "learning_rate": 0.0003871904484566103, | |
| "loss": 3.2957, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.753376804843967, | |
| "eval_accuracy": 0.37300792686416817, | |
| "eval_loss": 3.541806221008301, | |
| "eval_runtime": 82.3639, | |
| "eval_samples_per_second": 202.164, | |
| "eval_steps_per_second": 12.639, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.767931998136934, | |
| "grad_norm": 0.3416573703289032, | |
| "learning_rate": 0.0003870157251019219, | |
| "loss": 3.2733, | |
| "step": 61050 | |
| }, | |
| { | |
| "epoch": 17.7824871914299, | |
| "grad_norm": 0.37537774443626404, | |
| "learning_rate": 0.0003868410017472335, | |
| "loss": 3.28, | |
| "step": 61100 | |
| }, | |
| { | |
| "epoch": 17.797042384722868, | |
| "grad_norm": 0.3653329610824585, | |
| "learning_rate": 0.00038666627839254505, | |
| "loss": 3.2634, | |
| "step": 61150 | |
| }, | |
| { | |
| "epoch": 17.811597578015835, | |
| "grad_norm": 0.3708014190196991, | |
| "learning_rate": 0.0003864915550378567, | |
| "loss": 3.2737, | |
| "step": 61200 | |
| }, | |
| { | |
| "epoch": 17.826152771308802, | |
| "grad_norm": 0.36553969979286194, | |
| "learning_rate": 0.0003863168316831683, | |
| "loss": 3.2837, | |
| "step": 61250 | |
| }, | |
| { | |
| "epoch": 17.84070796460177, | |
| "grad_norm": 0.37409380078315735, | |
| "learning_rate": 0.0003861421083284799, | |
| "loss": 3.2819, | |
| "step": 61300 | |
| }, | |
| { | |
| "epoch": 17.855263157894736, | |
| "grad_norm": 0.36639204621315, | |
| "learning_rate": 0.0003859673849737914, | |
| "loss": 3.2827, | |
| "step": 61350 | |
| }, | |
| { | |
| "epoch": 17.869818351187703, | |
| "grad_norm": 0.3635062277317047, | |
| "learning_rate": 0.000385792661619103, | |
| "loss": 3.2733, | |
| "step": 61400 | |
| }, | |
| { | |
| "epoch": 17.88437354448067, | |
| "grad_norm": 0.3718494176864624, | |
| "learning_rate": 0.00038561793826441467, | |
| "loss": 3.2816, | |
| "step": 61450 | |
| }, | |
| { | |
| "epoch": 17.898928737773637, | |
| "grad_norm": 0.36888086795806885, | |
| "learning_rate": 0.00038544321490972626, | |
| "loss": 3.3043, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 17.913483931066605, | |
| "grad_norm": 0.4121055006980896, | |
| "learning_rate": 0.0003852684915550378, | |
| "loss": 3.286, | |
| "step": 61550 | |
| }, | |
| { | |
| "epoch": 17.92803912435957, | |
| "grad_norm": 0.3991578221321106, | |
| "learning_rate": 0.0003850937682003494, | |
| "loss": 3.2937, | |
| "step": 61600 | |
| }, | |
| { | |
| "epoch": 17.94259431765254, | |
| "grad_norm": 0.36824390292167664, | |
| "learning_rate": 0.000384919044845661, | |
| "loss": 3.2841, | |
| "step": 61650 | |
| }, | |
| { | |
| "epoch": 17.957149510945506, | |
| "grad_norm": 0.36976534128189087, | |
| "learning_rate": 0.0003847443214909726, | |
| "loss": 3.2958, | |
| "step": 61700 | |
| }, | |
| { | |
| "epoch": 17.971704704238473, | |
| "grad_norm": 0.3733980655670166, | |
| "learning_rate": 0.00038456959813628423, | |
| "loss": 3.2921, | |
| "step": 61750 | |
| }, | |
| { | |
| "epoch": 17.98625989753144, | |
| "grad_norm": 0.3652265965938568, | |
| "learning_rate": 0.0003843948747815958, | |
| "loss": 3.2806, | |
| "step": 61800 | |
| }, | |
| { | |
| "epoch": 18.000582207731718, | |
| "grad_norm": 0.4374643564224243, | |
| "learning_rate": 0.00038422015142690737, | |
| "loss": 3.2747, | |
| "step": 61850 | |
| }, | |
| { | |
| "epoch": 18.015137401024685, | |
| "grad_norm": 0.4197389781475067, | |
| "learning_rate": 0.00038404542807221896, | |
| "loss": 3.1641, | |
| "step": 61900 | |
| }, | |
| { | |
| "epoch": 18.029692594317652, | |
| "grad_norm": 0.37327665090560913, | |
| "learning_rate": 0.0003838707047175305, | |
| "loss": 3.186, | |
| "step": 61950 | |
| }, | |
| { | |
| "epoch": 18.04424778761062, | |
| "grad_norm": 0.3680726885795593, | |
| "learning_rate": 0.00038369598136284215, | |
| "loss": 3.196, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.04424778761062, | |
| "eval_accuracy": 0.3725329990229955, | |
| "eval_loss": 3.5523295402526855, | |
| "eval_runtime": 82.3873, | |
| "eval_samples_per_second": 202.106, | |
| "eval_steps_per_second": 12.635, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.058802980903586, | |
| "grad_norm": 0.36258381605148315, | |
| "learning_rate": 0.00038352125800815374, | |
| "loss": 3.1951, | |
| "step": 62050 | |
| }, | |
| { | |
| "epoch": 18.073358174196553, | |
| "grad_norm": 0.3668160140514374, | |
| "learning_rate": 0.00038334653465346534, | |
| "loss": 3.198, | |
| "step": 62100 | |
| }, | |
| { | |
| "epoch": 18.08791336748952, | |
| "grad_norm": 0.3587292432785034, | |
| "learning_rate": 0.0003831718112987769, | |
| "loss": 3.1863, | |
| "step": 62150 | |
| }, | |
| { | |
| "epoch": 18.102468560782487, | |
| "grad_norm": 0.3685011565685272, | |
| "learning_rate": 0.0003829970879440885, | |
| "loss": 3.1951, | |
| "step": 62200 | |
| }, | |
| { | |
| "epoch": 18.117023754075454, | |
| "grad_norm": 0.401020884513855, | |
| "learning_rate": 0.00038282236458940007, | |
| "loss": 3.2038, | |
| "step": 62250 | |
| }, | |
| { | |
| "epoch": 18.13157894736842, | |
| "grad_norm": 0.36973923444747925, | |
| "learning_rate": 0.0003826476412347117, | |
| "loss": 3.1994, | |
| "step": 62300 | |
| }, | |
| { | |
| "epoch": 18.14613414066139, | |
| "grad_norm": 0.3923777639865875, | |
| "learning_rate": 0.00038247291788002326, | |
| "loss": 3.2083, | |
| "step": 62350 | |
| }, | |
| { | |
| "epoch": 18.160689333954355, | |
| "grad_norm": 0.3689647614955902, | |
| "learning_rate": 0.00038229819452533485, | |
| "loss": 3.21, | |
| "step": 62400 | |
| }, | |
| { | |
| "epoch": 18.175244527247322, | |
| "grad_norm": 0.3853999078273773, | |
| "learning_rate": 0.00038212347117064644, | |
| "loss": 3.205, | |
| "step": 62450 | |
| }, | |
| { | |
| "epoch": 18.18979972054029, | |
| "grad_norm": 0.38615909218788147, | |
| "learning_rate": 0.000381948747815958, | |
| "loss": 3.2267, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 18.204354913833257, | |
| "grad_norm": 0.42501339316368103, | |
| "learning_rate": 0.0003817740244612696, | |
| "loss": 3.2173, | |
| "step": 62550 | |
| }, | |
| { | |
| "epoch": 18.218910107126224, | |
| "grad_norm": 0.3861982524394989, | |
| "learning_rate": 0.00038159930110658123, | |
| "loss": 3.2296, | |
| "step": 62600 | |
| }, | |
| { | |
| "epoch": 18.23346530041919, | |
| "grad_norm": 0.37562429904937744, | |
| "learning_rate": 0.0003814245777518928, | |
| "loss": 3.22, | |
| "step": 62650 | |
| }, | |
| { | |
| "epoch": 18.248020493712158, | |
| "grad_norm": 0.3911278247833252, | |
| "learning_rate": 0.0003812498543972044, | |
| "loss": 3.2266, | |
| "step": 62700 | |
| }, | |
| { | |
| "epoch": 18.262575687005125, | |
| "grad_norm": 0.36774829030036926, | |
| "learning_rate": 0.00038107513104251596, | |
| "loss": 3.2251, | |
| "step": 62750 | |
| }, | |
| { | |
| "epoch": 18.277130880298092, | |
| "grad_norm": 0.37904253602027893, | |
| "learning_rate": 0.00038090040768782755, | |
| "loss": 3.2255, | |
| "step": 62800 | |
| }, | |
| { | |
| "epoch": 18.29168607359106, | |
| "grad_norm": 0.38808226585388184, | |
| "learning_rate": 0.0003807256843331392, | |
| "loss": 3.2298, | |
| "step": 62850 | |
| }, | |
| { | |
| "epoch": 18.306241266884022, | |
| "grad_norm": 0.39811190962791443, | |
| "learning_rate": 0.0003805509609784508, | |
| "loss": 3.224, | |
| "step": 62900 | |
| }, | |
| { | |
| "epoch": 18.32079646017699, | |
| "grad_norm": 0.37187883257865906, | |
| "learning_rate": 0.00038037623762376233, | |
| "loss": 3.2344, | |
| "step": 62950 | |
| }, | |
| { | |
| "epoch": 18.335351653469957, | |
| "grad_norm": 0.3807899057865143, | |
| "learning_rate": 0.00038020151426907393, | |
| "loss": 3.2323, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.335351653469957, | |
| "eval_accuracy": 0.37293118153373367, | |
| "eval_loss": 3.550039768218994, | |
| "eval_runtime": 82.4172, | |
| "eval_samples_per_second": 202.033, | |
| "eval_steps_per_second": 12.631, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.349906846762924, | |
| "grad_norm": 0.4059674143791199, | |
| "learning_rate": 0.0003800267909143855, | |
| "loss": 3.2406, | |
| "step": 63050 | |
| }, | |
| { | |
| "epoch": 18.36446204005589, | |
| "grad_norm": 0.38597381114959717, | |
| "learning_rate": 0.00037985206755969706, | |
| "loss": 3.2429, | |
| "step": 63100 | |
| }, | |
| { | |
| "epoch": 18.379017233348858, | |
| "grad_norm": 0.3863556683063507, | |
| "learning_rate": 0.0003796773442050087, | |
| "loss": 3.2321, | |
| "step": 63150 | |
| }, | |
| { | |
| "epoch": 18.393572426641825, | |
| "grad_norm": 0.41818854212760925, | |
| "learning_rate": 0.0003795026208503203, | |
| "loss": 3.2439, | |
| "step": 63200 | |
| }, | |
| { | |
| "epoch": 18.408127619934792, | |
| "grad_norm": 0.3772805631160736, | |
| "learning_rate": 0.0003793278974956319, | |
| "loss": 3.2468, | |
| "step": 63250 | |
| }, | |
| { | |
| "epoch": 18.42268281322776, | |
| "grad_norm": 0.389891117811203, | |
| "learning_rate": 0.00037915317414094344, | |
| "loss": 3.2506, | |
| "step": 63300 | |
| }, | |
| { | |
| "epoch": 18.437238006520726, | |
| "grad_norm": 0.40573590993881226, | |
| "learning_rate": 0.00037897845078625503, | |
| "loss": 3.2588, | |
| "step": 63350 | |
| }, | |
| { | |
| "epoch": 18.451793199813693, | |
| "grad_norm": 0.3897237181663513, | |
| "learning_rate": 0.0003788037274315667, | |
| "loss": 3.2471, | |
| "step": 63400 | |
| }, | |
| { | |
| "epoch": 18.46634839310666, | |
| "grad_norm": 0.36086228489875793, | |
| "learning_rate": 0.0003786290040768783, | |
| "loss": 3.2595, | |
| "step": 63450 | |
| }, | |
| { | |
| "epoch": 18.480903586399627, | |
| "grad_norm": 0.39643728733062744, | |
| "learning_rate": 0.0003784542807221898, | |
| "loss": 3.2488, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 18.495458779692594, | |
| "grad_norm": 0.3505224287509918, | |
| "learning_rate": 0.0003782795573675014, | |
| "loss": 3.2464, | |
| "step": 63550 | |
| }, | |
| { | |
| "epoch": 18.51001397298556, | |
| "grad_norm": 0.38192400336265564, | |
| "learning_rate": 0.000378104834012813, | |
| "loss": 3.2547, | |
| "step": 63600 | |
| }, | |
| { | |
| "epoch": 18.52456916627853, | |
| "grad_norm": 0.4047112762928009, | |
| "learning_rate": 0.0003779301106581246, | |
| "loss": 3.249, | |
| "step": 63650 | |
| }, | |
| { | |
| "epoch": 18.539124359571495, | |
| "grad_norm": 0.381162166595459, | |
| "learning_rate": 0.0003777553873034362, | |
| "loss": 3.2482, | |
| "step": 63700 | |
| }, | |
| { | |
| "epoch": 18.553679552864462, | |
| "grad_norm": 0.4106951057910919, | |
| "learning_rate": 0.0003775806639487478, | |
| "loss": 3.2545, | |
| "step": 63750 | |
| }, | |
| { | |
| "epoch": 18.56823474615743, | |
| "grad_norm": 0.4451778531074524, | |
| "learning_rate": 0.0003774059405940594, | |
| "loss": 3.2494, | |
| "step": 63800 | |
| }, | |
| { | |
| "epoch": 18.582789939450397, | |
| "grad_norm": 0.36756718158721924, | |
| "learning_rate": 0.000377231217239371, | |
| "loss": 3.241, | |
| "step": 63850 | |
| }, | |
| { | |
| "epoch": 18.597345132743364, | |
| "grad_norm": 0.3792076110839844, | |
| "learning_rate": 0.0003770564938846825, | |
| "loss": 3.2537, | |
| "step": 63900 | |
| }, | |
| { | |
| "epoch": 18.61190032603633, | |
| "grad_norm": 0.40855872631073, | |
| "learning_rate": 0.00037688177052999416, | |
| "loss": 3.2643, | |
| "step": 63950 | |
| }, | |
| { | |
| "epoch": 18.626455519329298, | |
| "grad_norm": 0.37772053480148315, | |
| "learning_rate": 0.00037670704717530576, | |
| "loss": 3.2597, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.626455519329298, | |
| "eval_accuracy": 0.3732896398152424, | |
| "eval_loss": 3.5400493144989014, | |
| "eval_runtime": 82.4395, | |
| "eval_samples_per_second": 201.978, | |
| "eval_steps_per_second": 12.627, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.641010712622265, | |
| "grad_norm": 0.3802539110183716, | |
| "learning_rate": 0.00037653232382061735, | |
| "loss": 3.2536, | |
| "step": 64050 | |
| }, | |
| { | |
| "epoch": 18.655565905915232, | |
| "grad_norm": 0.38059988617897034, | |
| "learning_rate": 0.0003763576004659289, | |
| "loss": 3.2544, | |
| "step": 64100 | |
| }, | |
| { | |
| "epoch": 18.6701210992082, | |
| "grad_norm": 0.36282119154930115, | |
| "learning_rate": 0.0003761828771112405, | |
| "loss": 3.2496, | |
| "step": 64150 | |
| }, | |
| { | |
| "epoch": 18.684676292501166, | |
| "grad_norm": 0.41196170449256897, | |
| "learning_rate": 0.0003760081537565521, | |
| "loss": 3.2557, | |
| "step": 64200 | |
| }, | |
| { | |
| "epoch": 18.69923148579413, | |
| "grad_norm": 0.38784822821617126, | |
| "learning_rate": 0.00037583343040186373, | |
| "loss": 3.2566, | |
| "step": 64250 | |
| }, | |
| { | |
| "epoch": 18.713786679087097, | |
| "grad_norm": 0.3775848150253296, | |
| "learning_rate": 0.00037565870704717527, | |
| "loss": 3.2623, | |
| "step": 64300 | |
| }, | |
| { | |
| "epoch": 18.728341872380064, | |
| "grad_norm": 0.36481013894081116, | |
| "learning_rate": 0.00037548398369248687, | |
| "loss": 3.2446, | |
| "step": 64350 | |
| }, | |
| { | |
| "epoch": 18.74289706567303, | |
| "grad_norm": 0.3683626055717468, | |
| "learning_rate": 0.00037530926033779846, | |
| "loss": 3.2676, | |
| "step": 64400 | |
| }, | |
| { | |
| "epoch": 18.757452258965998, | |
| "grad_norm": 0.38007470965385437, | |
| "learning_rate": 0.00037513453698311, | |
| "loss": 3.2768, | |
| "step": 64450 | |
| }, | |
| { | |
| "epoch": 18.772007452258965, | |
| "grad_norm": 0.366379976272583, | |
| "learning_rate": 0.0003749598136284216, | |
| "loss": 3.2573, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 18.786562645551932, | |
| "grad_norm": 0.38161155581474304, | |
| "learning_rate": 0.00037478509027373324, | |
| "loss": 3.2514, | |
| "step": 64550 | |
| }, | |
| { | |
| "epoch": 18.8011178388449, | |
| "grad_norm": 0.39371827244758606, | |
| "learning_rate": 0.00037461036691904484, | |
| "loss": 3.2671, | |
| "step": 64600 | |
| }, | |
| { | |
| "epoch": 18.815673032137866, | |
| "grad_norm": 0.38536691665649414, | |
| "learning_rate": 0.0003744356435643564, | |
| "loss": 3.2749, | |
| "step": 64650 | |
| }, | |
| { | |
| "epoch": 18.830228225430833, | |
| "grad_norm": 0.3988786041736603, | |
| "learning_rate": 0.00037426092020966797, | |
| "loss": 3.2672, | |
| "step": 64700 | |
| }, | |
| { | |
| "epoch": 18.8447834187238, | |
| "grad_norm": 0.3931727707386017, | |
| "learning_rate": 0.00037408619685497957, | |
| "loss": 3.265, | |
| "step": 64750 | |
| }, | |
| { | |
| "epoch": 18.859338612016767, | |
| "grad_norm": 0.37959590554237366, | |
| "learning_rate": 0.0003739114735002912, | |
| "loss": 3.2739, | |
| "step": 64800 | |
| }, | |
| { | |
| "epoch": 18.873893805309734, | |
| "grad_norm": 0.35302025079727173, | |
| "learning_rate": 0.0003737367501456028, | |
| "loss": 3.2712, | |
| "step": 64850 | |
| }, | |
| { | |
| "epoch": 18.8884489986027, | |
| "grad_norm": 0.3610786497592926, | |
| "learning_rate": 0.00037356202679091435, | |
| "loss": 3.2664, | |
| "step": 64900 | |
| }, | |
| { | |
| "epoch": 18.90300419189567, | |
| "grad_norm": 0.3719325363636017, | |
| "learning_rate": 0.00037338730343622594, | |
| "loss": 3.2733, | |
| "step": 64950 | |
| }, | |
| { | |
| "epoch": 18.917559385188635, | |
| "grad_norm": 0.3856786787509918, | |
| "learning_rate": 0.00037321258008153754, | |
| "loss": 3.2769, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.917559385188635, | |
| "eval_accuracy": 0.3739452071248343, | |
| "eval_loss": 3.534527540206909, | |
| "eval_runtime": 82.2665, | |
| "eval_samples_per_second": 202.403, | |
| "eval_steps_per_second": 12.654, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.932114578481603, | |
| "grad_norm": 0.3829250633716583, | |
| "learning_rate": 0.0003730378567268491, | |
| "loss": 3.2727, | |
| "step": 65050 | |
| }, | |
| { | |
| "epoch": 18.94666977177457, | |
| "grad_norm": 0.406581848859787, | |
| "learning_rate": 0.0003728631333721607, | |
| "loss": 3.2615, | |
| "step": 65100 | |
| }, | |
| { | |
| "epoch": 18.961224965067537, | |
| "grad_norm": 0.3730480372905731, | |
| "learning_rate": 0.0003726884100174723, | |
| "loss": 3.2715, | |
| "step": 65150 | |
| }, | |
| { | |
| "epoch": 18.975780158360504, | |
| "grad_norm": 0.4020202159881592, | |
| "learning_rate": 0.0003725136866627839, | |
| "loss": 3.2747, | |
| "step": 65200 | |
| }, | |
| { | |
| "epoch": 18.99033535165347, | |
| "grad_norm": 0.3636406660079956, | |
| "learning_rate": 0.00037233896330809545, | |
| "loss": 3.2717, | |
| "step": 65250 | |
| }, | |
| { | |
| "epoch": 19.00465766185375, | |
| "grad_norm": 0.36889663338661194, | |
| "learning_rate": 0.00037216423995340705, | |
| "loss": 3.2464, | |
| "step": 65300 | |
| }, | |
| { | |
| "epoch": 19.019212855146716, | |
| "grad_norm": 0.3968113958835602, | |
| "learning_rate": 0.0003719895165987187, | |
| "loss": 3.1705, | |
| "step": 65350 | |
| }, | |
| { | |
| "epoch": 19.033768048439683, | |
| "grad_norm": 0.3783661723136902, | |
| "learning_rate": 0.0003718147932440303, | |
| "loss": 3.1824, | |
| "step": 65400 | |
| }, | |
| { | |
| "epoch": 19.04832324173265, | |
| "grad_norm": 0.38415125012397766, | |
| "learning_rate": 0.00037164006988934183, | |
| "loss": 3.1725, | |
| "step": 65450 | |
| }, | |
| { | |
| "epoch": 19.062878435025617, | |
| "grad_norm": 0.38789620995521545, | |
| "learning_rate": 0.0003714653465346534, | |
| "loss": 3.1882, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 19.077433628318584, | |
| "grad_norm": 0.37966445088386536, | |
| "learning_rate": 0.000371290623179965, | |
| "loss": 3.1833, | |
| "step": 65550 | |
| }, | |
| { | |
| "epoch": 19.09198882161155, | |
| "grad_norm": 0.3879067897796631, | |
| "learning_rate": 0.00037111589982527656, | |
| "loss": 3.1775, | |
| "step": 65600 | |
| }, | |
| { | |
| "epoch": 19.106544014904518, | |
| "grad_norm": 0.3996283710002899, | |
| "learning_rate": 0.0003709411764705882, | |
| "loss": 3.1802, | |
| "step": 65650 | |
| }, | |
| { | |
| "epoch": 19.121099208197485, | |
| "grad_norm": 0.41905948519706726, | |
| "learning_rate": 0.0003707664531158998, | |
| "loss": 3.1928, | |
| "step": 65700 | |
| }, | |
| { | |
| "epoch": 19.135654401490452, | |
| "grad_norm": 0.3763570785522461, | |
| "learning_rate": 0.0003705917297612114, | |
| "loss": 3.1936, | |
| "step": 65750 | |
| }, | |
| { | |
| "epoch": 19.15020959478342, | |
| "grad_norm": 0.4055962562561035, | |
| "learning_rate": 0.000370417006406523, | |
| "loss": 3.1886, | |
| "step": 65800 | |
| }, | |
| { | |
| "epoch": 19.164764788076386, | |
| "grad_norm": 0.3892934024333954, | |
| "learning_rate": 0.00037024228305183453, | |
| "loss": 3.1971, | |
| "step": 65850 | |
| }, | |
| { | |
| "epoch": 19.179319981369353, | |
| "grad_norm": 0.40365245938301086, | |
| "learning_rate": 0.0003700675596971461, | |
| "loss": 3.2076, | |
| "step": 65900 | |
| }, | |
| { | |
| "epoch": 19.19387517466232, | |
| "grad_norm": 0.38922300934791565, | |
| "learning_rate": 0.0003698928363424578, | |
| "loss": 3.2044, | |
| "step": 65950 | |
| }, | |
| { | |
| "epoch": 19.208430367955287, | |
| "grad_norm": 0.3984416425228119, | |
| "learning_rate": 0.00036971811298776937, | |
| "loss": 3.2138, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.208430367955287, | |
| "eval_accuracy": 0.3726682729515255, | |
| "eval_loss": 3.5525124073028564, | |
| "eval_runtime": 82.3134, | |
| "eval_samples_per_second": 202.288, | |
| "eval_steps_per_second": 12.647, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.222985561248255, | |
| "grad_norm": 0.4102519750595093, | |
| "learning_rate": 0.0003695433896330809, | |
| "loss": 3.2088, | |
| "step": 66050 | |
| }, | |
| { | |
| "epoch": 19.23754075454122, | |
| "grad_norm": 0.3882370591163635, | |
| "learning_rate": 0.0003693686662783925, | |
| "loss": 3.209, | |
| "step": 66100 | |
| }, | |
| { | |
| "epoch": 19.25209594783419, | |
| "grad_norm": 0.38305994868278503, | |
| "learning_rate": 0.0003691939429237041, | |
| "loss": 3.2105, | |
| "step": 66150 | |
| }, | |
| { | |
| "epoch": 19.266651141127156, | |
| "grad_norm": 0.395973265171051, | |
| "learning_rate": 0.00036901921956901575, | |
| "loss": 3.2212, | |
| "step": 66200 | |
| }, | |
| { | |
| "epoch": 19.281206334420123, | |
| "grad_norm": 0.3875166177749634, | |
| "learning_rate": 0.0003688444962143273, | |
| "loss": 3.2066, | |
| "step": 66250 | |
| }, | |
| { | |
| "epoch": 19.29576152771309, | |
| "grad_norm": 0.37727200984954834, | |
| "learning_rate": 0.0003686697728596389, | |
| "loss": 3.2248, | |
| "step": 66300 | |
| }, | |
| { | |
| "epoch": 19.310316721006053, | |
| "grad_norm": 0.3861169219017029, | |
| "learning_rate": 0.0003684950495049505, | |
| "loss": 3.2193, | |
| "step": 66350 | |
| }, | |
| { | |
| "epoch": 19.32487191429902, | |
| "grad_norm": 0.3775128424167633, | |
| "learning_rate": 0.000368320326150262, | |
| "loss": 3.2144, | |
| "step": 66400 | |
| }, | |
| { | |
| "epoch": 19.339427107591987, | |
| "grad_norm": 0.408060759305954, | |
| "learning_rate": 0.0003681456027955736, | |
| "loss": 3.2189, | |
| "step": 66450 | |
| }, | |
| { | |
| "epoch": 19.353982300884955, | |
| "grad_norm": 0.3870258331298828, | |
| "learning_rate": 0.00036797087944088526, | |
| "loss": 3.2189, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 19.36853749417792, | |
| "grad_norm": 0.38746318221092224, | |
| "learning_rate": 0.00036779615608619685, | |
| "loss": 3.2307, | |
| "step": 66550 | |
| }, | |
| { | |
| "epoch": 19.38309268747089, | |
| "grad_norm": 0.3949984908103943, | |
| "learning_rate": 0.0003676214327315084, | |
| "loss": 3.2303, | |
| "step": 66600 | |
| }, | |
| { | |
| "epoch": 19.397647880763856, | |
| "grad_norm": 0.3898114264011383, | |
| "learning_rate": 0.00036744670937682, | |
| "loss": 3.2279, | |
| "step": 66650 | |
| }, | |
| { | |
| "epoch": 19.412203074056823, | |
| "grad_norm": 0.38172972202301025, | |
| "learning_rate": 0.0003672719860221316, | |
| "loss": 3.2378, | |
| "step": 66700 | |
| }, | |
| { | |
| "epoch": 19.42675826734979, | |
| "grad_norm": 0.37363651394844055, | |
| "learning_rate": 0.00036709726266744323, | |
| "loss": 3.2337, | |
| "step": 66750 | |
| }, | |
| { | |
| "epoch": 19.441313460642757, | |
| "grad_norm": 0.40971848368644714, | |
| "learning_rate": 0.00036692253931275477, | |
| "loss": 3.2153, | |
| "step": 66800 | |
| }, | |
| { | |
| "epoch": 19.455868653935724, | |
| "grad_norm": 0.3720594644546509, | |
| "learning_rate": 0.00036674781595806636, | |
| "loss": 3.2321, | |
| "step": 66850 | |
| }, | |
| { | |
| "epoch": 19.47042384722869, | |
| "grad_norm": 0.4232275187969208, | |
| "learning_rate": 0.00036657309260337796, | |
| "loss": 3.2489, | |
| "step": 66900 | |
| }, | |
| { | |
| "epoch": 19.484979040521658, | |
| "grad_norm": 0.36863866448402405, | |
| "learning_rate": 0.00036639836924868955, | |
| "loss": 3.2389, | |
| "step": 66950 | |
| }, | |
| { | |
| "epoch": 19.499534233814625, | |
| "grad_norm": 0.4279008209705353, | |
| "learning_rate": 0.0003662236458940011, | |
| "loss": 3.2534, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.499534233814625, | |
| "eval_accuracy": 0.3729666747799683, | |
| "eval_loss": 3.5450198650360107, | |
| "eval_runtime": 82.2648, | |
| "eval_samples_per_second": 202.407, | |
| "eval_steps_per_second": 12.654, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.514089427107592, | |
| "grad_norm": 0.39464735984802246, | |
| "learning_rate": 0.00036604892253931274, | |
| "loss": 3.2533, | |
| "step": 67050 | |
| }, | |
| { | |
| "epoch": 19.52864462040056, | |
| "grad_norm": 0.3788803815841675, | |
| "learning_rate": 0.00036587419918462433, | |
| "loss": 3.2443, | |
| "step": 67100 | |
| }, | |
| { | |
| "epoch": 19.543199813693526, | |
| "grad_norm": 0.4050731956958771, | |
| "learning_rate": 0.00036569947582993593, | |
| "loss": 3.2492, | |
| "step": 67150 | |
| }, | |
| { | |
| "epoch": 19.557755006986493, | |
| "grad_norm": 0.38529258966445923, | |
| "learning_rate": 0.00036552475247524747, | |
| "loss": 3.2463, | |
| "step": 67200 | |
| }, | |
| { | |
| "epoch": 19.57231020027946, | |
| "grad_norm": 0.35914739966392517, | |
| "learning_rate": 0.00036535002912055906, | |
| "loss": 3.2354, | |
| "step": 67250 | |
| }, | |
| { | |
| "epoch": 19.586865393572428, | |
| "grad_norm": 0.40089380741119385, | |
| "learning_rate": 0.00036517530576587066, | |
| "loss": 3.244, | |
| "step": 67300 | |
| }, | |
| { | |
| "epoch": 19.601420586865395, | |
| "grad_norm": 0.3667324483394623, | |
| "learning_rate": 0.0003650005824111823, | |
| "loss": 3.2506, | |
| "step": 67350 | |
| }, | |
| { | |
| "epoch": 19.61597578015836, | |
| "grad_norm": 0.40627139806747437, | |
| "learning_rate": 0.00036482585905649385, | |
| "loss": 3.2479, | |
| "step": 67400 | |
| }, | |
| { | |
| "epoch": 19.63053097345133, | |
| "grad_norm": 0.3837718963623047, | |
| "learning_rate": 0.00036465113570180544, | |
| "loss": 3.2429, | |
| "step": 67450 | |
| }, | |
| { | |
| "epoch": 19.645086166744296, | |
| "grad_norm": 0.40455201268196106, | |
| "learning_rate": 0.00036447641234711703, | |
| "loss": 3.2465, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 19.659641360037263, | |
| "grad_norm": 0.39953795075416565, | |
| "learning_rate": 0.0003643016889924286, | |
| "loss": 3.2382, | |
| "step": 67550 | |
| }, | |
| { | |
| "epoch": 19.67419655333023, | |
| "grad_norm": 0.3991257846355438, | |
| "learning_rate": 0.0003641269656377402, | |
| "loss": 3.255, | |
| "step": 67600 | |
| }, | |
| { | |
| "epoch": 19.688751746623197, | |
| "grad_norm": 0.40475666522979736, | |
| "learning_rate": 0.0003639522422830518, | |
| "loss": 3.2422, | |
| "step": 67650 | |
| }, | |
| { | |
| "epoch": 19.70330693991616, | |
| "grad_norm": 0.4105340540409088, | |
| "learning_rate": 0.0003637775189283634, | |
| "loss": 3.2534, | |
| "step": 67700 | |
| }, | |
| { | |
| "epoch": 19.717862133209128, | |
| "grad_norm": 0.3834816813468933, | |
| "learning_rate": 0.00036360279557367495, | |
| "loss": 3.244, | |
| "step": 67750 | |
| }, | |
| { | |
| "epoch": 19.732417326502095, | |
| "grad_norm": 0.371440052986145, | |
| "learning_rate": 0.00036342807221898655, | |
| "loss": 3.2449, | |
| "step": 67800 | |
| }, | |
| { | |
| "epoch": 19.74697251979506, | |
| "grad_norm": 0.3959084749221802, | |
| "learning_rate": 0.00036325334886429814, | |
| "loss": 3.25, | |
| "step": 67850 | |
| }, | |
| { | |
| "epoch": 19.76152771308803, | |
| "grad_norm": 0.38034483790397644, | |
| "learning_rate": 0.0003630786255096098, | |
| "loss": 3.2492, | |
| "step": 67900 | |
| }, | |
| { | |
| "epoch": 19.776082906380996, | |
| "grad_norm": 0.3578837513923645, | |
| "learning_rate": 0.00036290390215492133, | |
| "loss": 3.2546, | |
| "step": 67950 | |
| }, | |
| { | |
| "epoch": 19.790638099673963, | |
| "grad_norm": 0.3825294077396393, | |
| "learning_rate": 0.0003627291788002329, | |
| "loss": 3.245, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.790638099673963, | |
| "eval_accuracy": 0.37374705608790854, | |
| "eval_loss": 3.5345418453216553, | |
| "eval_runtime": 82.3426, | |
| "eval_samples_per_second": 202.216, | |
| "eval_steps_per_second": 12.642, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.80519329296693, | |
| "grad_norm": 0.37749066948890686, | |
| "learning_rate": 0.0003625544554455445, | |
| "loss": 3.2551, | |
| "step": 68050 | |
| }, | |
| { | |
| "epoch": 19.819748486259897, | |
| "grad_norm": 0.3952888548374176, | |
| "learning_rate": 0.0003623797320908561, | |
| "loss": 3.2504, | |
| "step": 68100 | |
| }, | |
| { | |
| "epoch": 19.834303679552864, | |
| "grad_norm": 0.4495806396007538, | |
| "learning_rate": 0.00036220500873616776, | |
| "loss": 3.265, | |
| "step": 68150 | |
| }, | |
| { | |
| "epoch": 19.84885887284583, | |
| "grad_norm": 0.3956485688686371, | |
| "learning_rate": 0.0003620302853814793, | |
| "loss": 3.2536, | |
| "step": 68200 | |
| }, | |
| { | |
| "epoch": 19.863414066138798, | |
| "grad_norm": 0.3715929687023163, | |
| "learning_rate": 0.0003618555620267909, | |
| "loss": 3.2611, | |
| "step": 68250 | |
| }, | |
| { | |
| "epoch": 19.877969259431765, | |
| "grad_norm": 0.40561985969543457, | |
| "learning_rate": 0.0003616808386721025, | |
| "loss": 3.2502, | |
| "step": 68300 | |
| }, | |
| { | |
| "epoch": 19.892524452724732, | |
| "grad_norm": 0.37451308965682983, | |
| "learning_rate": 0.00036150611531741403, | |
| "loss": 3.2545, | |
| "step": 68350 | |
| }, | |
| { | |
| "epoch": 19.9070796460177, | |
| "grad_norm": 0.36757317185401917, | |
| "learning_rate": 0.0003613313919627256, | |
| "loss": 3.2601, | |
| "step": 68400 | |
| }, | |
| { | |
| "epoch": 19.921634839310666, | |
| "grad_norm": 0.3995347023010254, | |
| "learning_rate": 0.00036115666860803727, | |
| "loss": 3.266, | |
| "step": 68450 | |
| }, | |
| { | |
| "epoch": 19.936190032603633, | |
| "grad_norm": 0.39765140414237976, | |
| "learning_rate": 0.00036098194525334887, | |
| "loss": 3.2672, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 19.9507452258966, | |
| "grad_norm": 0.3967185914516449, | |
| "learning_rate": 0.0003608072218986604, | |
| "loss": 3.2588, | |
| "step": 68550 | |
| }, | |
| { | |
| "epoch": 19.965300419189568, | |
| "grad_norm": 0.40620025992393494, | |
| "learning_rate": 0.000360632498543972, | |
| "loss": 3.2689, | |
| "step": 68600 | |
| }, | |
| { | |
| "epoch": 19.979855612482535, | |
| "grad_norm": 0.39762723445892334, | |
| "learning_rate": 0.0003604577751892836, | |
| "loss": 3.2654, | |
| "step": 68650 | |
| }, | |
| { | |
| "epoch": 19.9944108057755, | |
| "grad_norm": 0.38677430152893066, | |
| "learning_rate": 0.00036028305183459513, | |
| "loss": 3.266, | |
| "step": 68700 | |
| }, | |
| { | |
| "epoch": 20.00873311597578, | |
| "grad_norm": 0.49194103479385376, | |
| "learning_rate": 0.0003601083284799068, | |
| "loss": 3.1981, | |
| "step": 68750 | |
| }, | |
| { | |
| "epoch": 20.023288309268747, | |
| "grad_norm": 0.3924783170223236, | |
| "learning_rate": 0.0003599336051252184, | |
| "loss": 3.1548, | |
| "step": 68800 | |
| }, | |
| { | |
| "epoch": 20.037843502561714, | |
| "grad_norm": 0.3935804069042206, | |
| "learning_rate": 0.00035975888177052997, | |
| "loss": 3.1611, | |
| "step": 68850 | |
| }, | |
| { | |
| "epoch": 20.05239869585468, | |
| "grad_norm": 0.38737156987190247, | |
| "learning_rate": 0.0003595841584158415, | |
| "loss": 3.1656, | |
| "step": 68900 | |
| }, | |
| { | |
| "epoch": 20.066953889147648, | |
| "grad_norm": 0.41584619879722595, | |
| "learning_rate": 0.0003594094350611531, | |
| "loss": 3.1792, | |
| "step": 68950 | |
| }, | |
| { | |
| "epoch": 20.081509082440615, | |
| "grad_norm": 0.3923519551753998, | |
| "learning_rate": 0.00035923471170646475, | |
| "loss": 3.1755, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.081509082440615, | |
| "eval_accuracy": 0.3730060464272816, | |
| "eval_loss": 3.55251145362854, | |
| "eval_runtime": 82.7346, | |
| "eval_samples_per_second": 201.258, | |
| "eval_steps_per_second": 12.582, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.096064275733582, | |
| "grad_norm": 0.403720885515213, | |
| "learning_rate": 0.00035905998835177635, | |
| "loss": 3.1676, | |
| "step": 69050 | |
| }, | |
| { | |
| "epoch": 20.11061946902655, | |
| "grad_norm": 0.38353613018989563, | |
| "learning_rate": 0.00035888526499708794, | |
| "loss": 3.1795, | |
| "step": 69100 | |
| }, | |
| { | |
| "epoch": 20.125174662319516, | |
| "grad_norm": 0.4610157907009125, | |
| "learning_rate": 0.0003587105416423995, | |
| "loss": 3.1753, | |
| "step": 69150 | |
| }, | |
| { | |
| "epoch": 20.139729855612483, | |
| "grad_norm": 0.41082763671875, | |
| "learning_rate": 0.0003585358182877111, | |
| "loss": 3.1827, | |
| "step": 69200 | |
| }, | |
| { | |
| "epoch": 20.15428504890545, | |
| "grad_norm": 0.3844628632068634, | |
| "learning_rate": 0.00035836109493302267, | |
| "loss": 3.1797, | |
| "step": 69250 | |
| }, | |
| { | |
| "epoch": 20.168840242198417, | |
| "grad_norm": 0.4087962210178375, | |
| "learning_rate": 0.0003581863715783343, | |
| "loss": 3.1975, | |
| "step": 69300 | |
| }, | |
| { | |
| "epoch": 20.183395435491384, | |
| "grad_norm": 0.399774432182312, | |
| "learning_rate": 0.00035801164822364586, | |
| "loss": 3.186, | |
| "step": 69350 | |
| }, | |
| { | |
| "epoch": 20.19795062878435, | |
| "grad_norm": 0.4322206676006317, | |
| "learning_rate": 0.00035783692486895745, | |
| "loss": 3.1824, | |
| "step": 69400 | |
| }, | |
| { | |
| "epoch": 20.21250582207732, | |
| "grad_norm": 0.3971948027610779, | |
| "learning_rate": 0.00035766220151426905, | |
| "loss": 3.2043, | |
| "step": 69450 | |
| }, | |
| { | |
| "epoch": 20.227061015370285, | |
| "grad_norm": 0.40004590153694153, | |
| "learning_rate": 0.0003574874781595806, | |
| "loss": 3.2011, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 20.241616208663253, | |
| "grad_norm": 0.4095022976398468, | |
| "learning_rate": 0.00035731275480489224, | |
| "loss": 3.2094, | |
| "step": 69550 | |
| }, | |
| { | |
| "epoch": 20.25617140195622, | |
| "grad_norm": 0.3898414671421051, | |
| "learning_rate": 0.00035713803145020383, | |
| "loss": 3.2054, | |
| "step": 69600 | |
| }, | |
| { | |
| "epoch": 20.270726595249187, | |
| "grad_norm": 0.37962237000465393, | |
| "learning_rate": 0.0003569633080955154, | |
| "loss": 3.203, | |
| "step": 69650 | |
| }, | |
| { | |
| "epoch": 20.28528178854215, | |
| "grad_norm": 0.38528409600257874, | |
| "learning_rate": 0.00035678858474082697, | |
| "loss": 3.2061, | |
| "step": 69700 | |
| }, | |
| { | |
| "epoch": 20.299836981835117, | |
| "grad_norm": 0.3884774446487427, | |
| "learning_rate": 0.00035661386138613856, | |
| "loss": 3.2047, | |
| "step": 69750 | |
| }, | |
| { | |
| "epoch": 20.314392175128084, | |
| "grad_norm": 0.4096026122570038, | |
| "learning_rate": 0.00035643913803145015, | |
| "loss": 3.2106, | |
| "step": 69800 | |
| }, | |
| { | |
| "epoch": 20.32894736842105, | |
| "grad_norm": 0.40751922130584717, | |
| "learning_rate": 0.0003562644146767618, | |
| "loss": 3.218, | |
| "step": 69850 | |
| }, | |
| { | |
| "epoch": 20.34350256171402, | |
| "grad_norm": 0.3940451145172119, | |
| "learning_rate": 0.00035608969132207334, | |
| "loss": 3.2141, | |
| "step": 69900 | |
| }, | |
| { | |
| "epoch": 20.358057755006985, | |
| "grad_norm": 0.3952007591724396, | |
| "learning_rate": 0.00035591496796738494, | |
| "loss": 3.2117, | |
| "step": 69950 | |
| }, | |
| { | |
| "epoch": 20.372612948299953, | |
| "grad_norm": 0.40748631954193115, | |
| "learning_rate": 0.00035574024461269653, | |
| "loss": 3.2154, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.372612948299953, | |
| "eval_accuracy": 0.37326613435415984, | |
| "eval_loss": 3.546116590499878, | |
| "eval_runtime": 82.7749, | |
| "eval_samples_per_second": 201.16, | |
| "eval_steps_per_second": 12.576, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.38716814159292, | |
| "grad_norm": 0.39818665385246277, | |
| "learning_rate": 0.0003555655212580081, | |
| "loss": 3.2234, | |
| "step": 70050 | |
| }, | |
| { | |
| "epoch": 20.401723334885887, | |
| "grad_norm": 0.41679659485816956, | |
| "learning_rate": 0.00035539079790331967, | |
| "loss": 3.2246, | |
| "step": 70100 | |
| }, | |
| { | |
| "epoch": 20.416278528178854, | |
| "grad_norm": 0.43143290281295776, | |
| "learning_rate": 0.0003552160745486313, | |
| "loss": 3.2327, | |
| "step": 70150 | |
| }, | |
| { | |
| "epoch": 20.43083372147182, | |
| "grad_norm": 0.4188302457332611, | |
| "learning_rate": 0.0003550413511939429, | |
| "loss": 3.2263, | |
| "step": 70200 | |
| }, | |
| { | |
| "epoch": 20.445388914764788, | |
| "grad_norm": 0.3997771739959717, | |
| "learning_rate": 0.0003548666278392545, | |
| "loss": 3.2126, | |
| "step": 70250 | |
| }, | |
| { | |
| "epoch": 20.459944108057755, | |
| "grad_norm": 0.42035701870918274, | |
| "learning_rate": 0.00035469190448456604, | |
| "loss": 3.2139, | |
| "step": 70300 | |
| }, | |
| { | |
| "epoch": 20.474499301350722, | |
| "grad_norm": 0.4001785218715668, | |
| "learning_rate": 0.00035451718112987764, | |
| "loss": 3.2303, | |
| "step": 70350 | |
| }, | |
| { | |
| "epoch": 20.48905449464369, | |
| "grad_norm": 0.3872750401496887, | |
| "learning_rate": 0.0003543424577751893, | |
| "loss": 3.2224, | |
| "step": 70400 | |
| }, | |
| { | |
| "epoch": 20.503609687936656, | |
| "grad_norm": 0.38086098432540894, | |
| "learning_rate": 0.0003541677344205009, | |
| "loss": 3.2276, | |
| "step": 70450 | |
| }, | |
| { | |
| "epoch": 20.518164881229623, | |
| "grad_norm": 0.40968039631843567, | |
| "learning_rate": 0.0003539930110658124, | |
| "loss": 3.2238, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 20.53272007452259, | |
| "grad_norm": 0.41206908226013184, | |
| "learning_rate": 0.000353818287711124, | |
| "loss": 3.235, | |
| "step": 70550 | |
| }, | |
| { | |
| "epoch": 20.547275267815557, | |
| "grad_norm": 0.3902190923690796, | |
| "learning_rate": 0.0003536435643564356, | |
| "loss": 3.247, | |
| "step": 70600 | |
| }, | |
| { | |
| "epoch": 20.561830461108524, | |
| "grad_norm": 0.3957805931568146, | |
| "learning_rate": 0.00035346884100174715, | |
| "loss": 3.2319, | |
| "step": 70650 | |
| }, | |
| { | |
| "epoch": 20.57638565440149, | |
| "grad_norm": 0.3716483414173126, | |
| "learning_rate": 0.0003532941176470588, | |
| "loss": 3.2338, | |
| "step": 70700 | |
| }, | |
| { | |
| "epoch": 20.59094084769446, | |
| "grad_norm": 0.38117969036102295, | |
| "learning_rate": 0.0003531193942923704, | |
| "loss": 3.2346, | |
| "step": 70750 | |
| }, | |
| { | |
| "epoch": 20.605496040987425, | |
| "grad_norm": 0.4173533320426941, | |
| "learning_rate": 0.000352944670937682, | |
| "loss": 3.2255, | |
| "step": 70800 | |
| }, | |
| { | |
| "epoch": 20.620051234280393, | |
| "grad_norm": 0.3798835277557373, | |
| "learning_rate": 0.0003527699475829935, | |
| "loss": 3.2407, | |
| "step": 70850 | |
| }, | |
| { | |
| "epoch": 20.63460642757336, | |
| "grad_norm": 0.5332134366035461, | |
| "learning_rate": 0.0003525952242283051, | |
| "loss": 3.2254, | |
| "step": 70900 | |
| }, | |
| { | |
| "epoch": 20.649161620866327, | |
| "grad_norm": 0.3963099718093872, | |
| "learning_rate": 0.00035242050087361677, | |
| "loss": 3.2358, | |
| "step": 70950 | |
| }, | |
| { | |
| "epoch": 20.663716814159294, | |
| "grad_norm": 0.4075072407722473, | |
| "learning_rate": 0.00035224577751892836, | |
| "loss": 3.2398, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.663716814159294, | |
| "eval_accuracy": 0.37394649992519385, | |
| "eval_loss": 3.538309097290039, | |
| "eval_runtime": 82.8012, | |
| "eval_samples_per_second": 201.096, | |
| "eval_steps_per_second": 12.572, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.678272007452257, | |
| "grad_norm": 0.39389562606811523, | |
| "learning_rate": 0.0003520710541642399, | |
| "loss": 3.2375, | |
| "step": 71050 | |
| }, | |
| { | |
| "epoch": 20.692827200745224, | |
| "grad_norm": 0.4054398834705353, | |
| "learning_rate": 0.0003518963308095515, | |
| "loss": 3.2415, | |
| "step": 71100 | |
| }, | |
| { | |
| "epoch": 20.70738239403819, | |
| "grad_norm": 0.4022291600704193, | |
| "learning_rate": 0.0003517216074548631, | |
| "loss": 3.2343, | |
| "step": 71150 | |
| }, | |
| { | |
| "epoch": 20.72193758733116, | |
| "grad_norm": 0.40785783529281616, | |
| "learning_rate": 0.0003515468841001747, | |
| "loss": 3.226, | |
| "step": 71200 | |
| }, | |
| { | |
| "epoch": 20.736492780624125, | |
| "grad_norm": 0.429970383644104, | |
| "learning_rate": 0.00035137216074548634, | |
| "loss": 3.2488, | |
| "step": 71250 | |
| }, | |
| { | |
| "epoch": 20.751047973917093, | |
| "grad_norm": 0.38982483744621277, | |
| "learning_rate": 0.0003511974373907979, | |
| "loss": 3.2345, | |
| "step": 71300 | |
| }, | |
| { | |
| "epoch": 20.76560316721006, | |
| "grad_norm": 0.3958439230918884, | |
| "learning_rate": 0.00035102271403610947, | |
| "loss": 3.2453, | |
| "step": 71350 | |
| }, | |
| { | |
| "epoch": 20.780158360503027, | |
| "grad_norm": 0.4144258201122284, | |
| "learning_rate": 0.00035084799068142106, | |
| "loss": 3.2378, | |
| "step": 71400 | |
| }, | |
| { | |
| "epoch": 20.794713553795994, | |
| "grad_norm": 0.39274704456329346, | |
| "learning_rate": 0.0003506732673267326, | |
| "loss": 3.2344, | |
| "step": 71450 | |
| }, | |
| { | |
| "epoch": 20.80926874708896, | |
| "grad_norm": 0.4060254395008087, | |
| "learning_rate": 0.00035049854397204425, | |
| "loss": 3.2349, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 20.823823940381928, | |
| "grad_norm": 0.3565615713596344, | |
| "learning_rate": 0.00035032382061735585, | |
| "loss": 3.2591, | |
| "step": 71550 | |
| }, | |
| { | |
| "epoch": 20.838379133674895, | |
| "grad_norm": 0.40499451756477356, | |
| "learning_rate": 0.00035014909726266744, | |
| "loss": 3.2472, | |
| "step": 71600 | |
| }, | |
| { | |
| "epoch": 20.852934326967862, | |
| "grad_norm": 0.38859230279922485, | |
| "learning_rate": 0.000349974373907979, | |
| "loss": 3.2547, | |
| "step": 71650 | |
| }, | |
| { | |
| "epoch": 20.86748952026083, | |
| "grad_norm": 0.404277503490448, | |
| "learning_rate": 0.0003497996505532906, | |
| "loss": 3.2375, | |
| "step": 71700 | |
| }, | |
| { | |
| "epoch": 20.882044713553796, | |
| "grad_norm": 0.39116883277893066, | |
| "learning_rate": 0.00034962492719860217, | |
| "loss": 3.2575, | |
| "step": 71750 | |
| }, | |
| { | |
| "epoch": 20.896599906846763, | |
| "grad_norm": 0.4242528975009918, | |
| "learning_rate": 0.0003494502038439138, | |
| "loss": 3.248, | |
| "step": 71800 | |
| }, | |
| { | |
| "epoch": 20.91115510013973, | |
| "grad_norm": 0.7736283540725708, | |
| "learning_rate": 0.00034927548048922536, | |
| "loss": 3.257, | |
| "step": 71850 | |
| }, | |
| { | |
| "epoch": 20.925710293432697, | |
| "grad_norm": 0.3877331018447876, | |
| "learning_rate": 0.00034910075713453695, | |
| "loss": 3.2462, | |
| "step": 71900 | |
| }, | |
| { | |
| "epoch": 20.940265486725664, | |
| "grad_norm": 0.3922559916973114, | |
| "learning_rate": 0.00034892603377984855, | |
| "loss": 3.2517, | |
| "step": 71950 | |
| }, | |
| { | |
| "epoch": 20.95482068001863, | |
| "grad_norm": 0.38187119364738464, | |
| "learning_rate": 0.0003487513104251601, | |
| "loss": 3.2521, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.95482068001863, | |
| "eval_accuracy": 0.37429156009388553, | |
| "eval_loss": 3.5316383838653564, | |
| "eval_runtime": 82.7894, | |
| "eval_samples_per_second": 201.125, | |
| "eval_steps_per_second": 12.574, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.9693758733116, | |
| "grad_norm": 0.39174169301986694, | |
| "learning_rate": 0.0003485765870704717, | |
| "loss": 3.2583, | |
| "step": 72050 | |
| }, | |
| { | |
| "epoch": 20.983931066604566, | |
| "grad_norm": 0.39717811346054077, | |
| "learning_rate": 0.00034840186371578333, | |
| "loss": 3.2591, | |
| "step": 72100 | |
| }, | |
| { | |
| "epoch": 20.998486259897533, | |
| "grad_norm": 0.40414169430732727, | |
| "learning_rate": 0.0003482271403610949, | |
| "loss": 3.2418, | |
| "step": 72150 | |
| }, | |
| { | |
| "epoch": 21.01280857009781, | |
| "grad_norm": 0.38366419076919556, | |
| "learning_rate": 0.0003480524170064065, | |
| "loss": 3.1705, | |
| "step": 72200 | |
| }, | |
| { | |
| "epoch": 21.027363763390778, | |
| "grad_norm": 0.4076693058013916, | |
| "learning_rate": 0.00034787769365171806, | |
| "loss": 3.1432, | |
| "step": 72250 | |
| }, | |
| { | |
| "epoch": 21.041918956683745, | |
| "grad_norm": 0.41582274436950684, | |
| "learning_rate": 0.00034770297029702965, | |
| "loss": 3.1508, | |
| "step": 72300 | |
| }, | |
| { | |
| "epoch": 21.05647414997671, | |
| "grad_norm": 0.408663809299469, | |
| "learning_rate": 0.0003475282469423413, | |
| "loss": 3.1534, | |
| "step": 72350 | |
| }, | |
| { | |
| "epoch": 21.07102934326968, | |
| "grad_norm": 0.4284273684024811, | |
| "learning_rate": 0.0003473535235876529, | |
| "loss": 3.1625, | |
| "step": 72400 | |
| }, | |
| { | |
| "epoch": 21.085584536562646, | |
| "grad_norm": 0.427682489156723, | |
| "learning_rate": 0.00034717880023296444, | |
| "loss": 3.1709, | |
| "step": 72450 | |
| }, | |
| { | |
| "epoch": 21.100139729855613, | |
| "grad_norm": 0.3825227916240692, | |
| "learning_rate": 0.00034700407687827603, | |
| "loss": 3.1707, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 21.11469492314858, | |
| "grad_norm": 0.4480525851249695, | |
| "learning_rate": 0.0003468293535235876, | |
| "loss": 3.173, | |
| "step": 72550 | |
| }, | |
| { | |
| "epoch": 21.129250116441547, | |
| "grad_norm": 0.40936607122421265, | |
| "learning_rate": 0.00034665463016889916, | |
| "loss": 3.1757, | |
| "step": 72600 | |
| }, | |
| { | |
| "epoch": 21.143805309734514, | |
| "grad_norm": 0.40996667742729187, | |
| "learning_rate": 0.0003464799068142108, | |
| "loss": 3.1762, | |
| "step": 72650 | |
| }, | |
| { | |
| "epoch": 21.15836050302748, | |
| "grad_norm": 0.38334524631500244, | |
| "learning_rate": 0.0003463051834595224, | |
| "loss": 3.1738, | |
| "step": 72700 | |
| }, | |
| { | |
| "epoch": 21.172915696320448, | |
| "grad_norm": 0.38095003366470337, | |
| "learning_rate": 0.000346130460104834, | |
| "loss": 3.1707, | |
| "step": 72750 | |
| }, | |
| { | |
| "epoch": 21.187470889613415, | |
| "grad_norm": 0.37812045216560364, | |
| "learning_rate": 0.00034595573675014554, | |
| "loss": 3.1876, | |
| "step": 72800 | |
| }, | |
| { | |
| "epoch": 21.202026082906382, | |
| "grad_norm": 0.4450712502002716, | |
| "learning_rate": 0.00034578101339545714, | |
| "loss": 3.1864, | |
| "step": 72850 | |
| }, | |
| { | |
| "epoch": 21.21658127619935, | |
| "grad_norm": 0.41464632749557495, | |
| "learning_rate": 0.0003456062900407688, | |
| "loss": 3.1856, | |
| "step": 72900 | |
| }, | |
| { | |
| "epoch": 21.231136469492316, | |
| "grad_norm": 0.4131869971752167, | |
| "learning_rate": 0.0003454315666860804, | |
| "loss": 3.2029, | |
| "step": 72950 | |
| }, | |
| { | |
| "epoch": 21.245691662785283, | |
| "grad_norm": 0.41746553778648376, | |
| "learning_rate": 0.0003452568433313919, | |
| "loss": 3.1916, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.245691662785283, | |
| "eval_accuracy": 0.37350824060330995, | |
| "eval_loss": 3.5468664169311523, | |
| "eval_runtime": 83.0316, | |
| "eval_samples_per_second": 200.538, | |
| "eval_steps_per_second": 12.537, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.260246856078247, | |
| "grad_norm": 0.4272761344909668, | |
| "learning_rate": 0.0003450821199767035, | |
| "loss": 3.1957, | |
| "step": 73050 | |
| }, | |
| { | |
| "epoch": 21.274802049371214, | |
| "grad_norm": 0.440285861492157, | |
| "learning_rate": 0.0003449073966220151, | |
| "loss": 3.2049, | |
| "step": 73100 | |
| }, | |
| { | |
| "epoch": 21.28935724266418, | |
| "grad_norm": 0.4013405442237854, | |
| "learning_rate": 0.0003447326732673267, | |
| "loss": 3.182, | |
| "step": 73150 | |
| }, | |
| { | |
| "epoch": 21.303912435957148, | |
| "grad_norm": 0.37748634815216064, | |
| "learning_rate": 0.0003445579499126383, | |
| "loss": 3.2073, | |
| "step": 73200 | |
| }, | |
| { | |
| "epoch": 21.318467629250115, | |
| "grad_norm": 0.3896017074584961, | |
| "learning_rate": 0.0003443832265579499, | |
| "loss": 3.1979, | |
| "step": 73250 | |
| }, | |
| { | |
| "epoch": 21.333022822543082, | |
| "grad_norm": 0.4161280691623688, | |
| "learning_rate": 0.0003442085032032615, | |
| "loss": 3.2106, | |
| "step": 73300 | |
| }, | |
| { | |
| "epoch": 21.34757801583605, | |
| "grad_norm": 0.3918216824531555, | |
| "learning_rate": 0.0003440337798485731, | |
| "loss": 3.201, | |
| "step": 73350 | |
| }, | |
| { | |
| "epoch": 21.362133209129016, | |
| "grad_norm": 0.39482593536376953, | |
| "learning_rate": 0.0003438590564938846, | |
| "loss": 3.2021, | |
| "step": 73400 | |
| }, | |
| { | |
| "epoch": 21.376688402421983, | |
| "grad_norm": 0.37131577730178833, | |
| "learning_rate": 0.0003436843331391962, | |
| "loss": 3.1917, | |
| "step": 73450 | |
| }, | |
| { | |
| "epoch": 21.39124359571495, | |
| "grad_norm": 0.4280555248260498, | |
| "learning_rate": 0.00034350960978450786, | |
| "loss": 3.2001, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 21.405798789007918, | |
| "grad_norm": 0.4589449465274811, | |
| "learning_rate": 0.00034333488642981946, | |
| "loss": 3.2, | |
| "step": 73550 | |
| }, | |
| { | |
| "epoch": 21.420353982300885, | |
| "grad_norm": 0.4011155068874359, | |
| "learning_rate": 0.000343160163075131, | |
| "loss": 3.2118, | |
| "step": 73600 | |
| }, | |
| { | |
| "epoch": 21.43490917559385, | |
| "grad_norm": 0.432386577129364, | |
| "learning_rate": 0.0003429854397204426, | |
| "loss": 3.2238, | |
| "step": 73650 | |
| }, | |
| { | |
| "epoch": 21.44946436888682, | |
| "grad_norm": 0.4160323441028595, | |
| "learning_rate": 0.0003428107163657542, | |
| "loss": 3.209, | |
| "step": 73700 | |
| }, | |
| { | |
| "epoch": 21.464019562179786, | |
| "grad_norm": 0.40614748001098633, | |
| "learning_rate": 0.00034263599301106583, | |
| "loss": 3.2047, | |
| "step": 73750 | |
| }, | |
| { | |
| "epoch": 21.478574755472753, | |
| "grad_norm": 0.39390140771865845, | |
| "learning_rate": 0.0003424612696563774, | |
| "loss": 3.2102, | |
| "step": 73800 | |
| }, | |
| { | |
| "epoch": 21.49312994876572, | |
| "grad_norm": 0.4248008131980896, | |
| "learning_rate": 0.00034228654630168897, | |
| "loss": 3.216, | |
| "step": 73850 | |
| }, | |
| { | |
| "epoch": 21.507685142058687, | |
| "grad_norm": 0.4202058017253876, | |
| "learning_rate": 0.00034211182294700056, | |
| "loss": 3.2114, | |
| "step": 73900 | |
| }, | |
| { | |
| "epoch": 21.522240335351654, | |
| "grad_norm": 0.3771965801715851, | |
| "learning_rate": 0.0003419370995923121, | |
| "loss": 3.222, | |
| "step": 73950 | |
| }, | |
| { | |
| "epoch": 21.53679552864462, | |
| "grad_norm": 0.399825781583786, | |
| "learning_rate": 0.0003417623762376237, | |
| "loss": 3.2129, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.53679552864462, | |
| "eval_accuracy": 0.37412596412055904, | |
| "eval_loss": 3.5408825874328613, | |
| "eval_runtime": 82.9334, | |
| "eval_samples_per_second": 200.776, | |
| "eval_steps_per_second": 12.552, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.551350721937588, | |
| "grad_norm": 0.4415282905101776, | |
| "learning_rate": 0.00034158765288293534, | |
| "loss": 3.2238, | |
| "step": 74050 | |
| }, | |
| { | |
| "epoch": 21.565905915230555, | |
| "grad_norm": 0.40851542353630066, | |
| "learning_rate": 0.00034141292952824694, | |
| "loss": 3.2232, | |
| "step": 74100 | |
| }, | |
| { | |
| "epoch": 21.580461108523522, | |
| "grad_norm": 0.3917493224143982, | |
| "learning_rate": 0.0003412382061735585, | |
| "loss": 3.2327, | |
| "step": 74150 | |
| }, | |
| { | |
| "epoch": 21.59501630181649, | |
| "grad_norm": 0.42648059129714966, | |
| "learning_rate": 0.0003410634828188701, | |
| "loss": 3.2212, | |
| "step": 74200 | |
| }, | |
| { | |
| "epoch": 21.609571495109456, | |
| "grad_norm": 0.38889065384864807, | |
| "learning_rate": 0.00034088875946418167, | |
| "loss": 3.2238, | |
| "step": 74250 | |
| }, | |
| { | |
| "epoch": 21.624126688402423, | |
| "grad_norm": 0.40096256136894226, | |
| "learning_rate": 0.0003407140361094933, | |
| "loss": 3.2296, | |
| "step": 74300 | |
| }, | |
| { | |
| "epoch": 21.63868188169539, | |
| "grad_norm": 0.393991082906723, | |
| "learning_rate": 0.00034053931275480486, | |
| "loss": 3.2191, | |
| "step": 74350 | |
| }, | |
| { | |
| "epoch": 21.653237074988354, | |
| "grad_norm": 0.39421722292900085, | |
| "learning_rate": 0.00034036458940011645, | |
| "loss": 3.2302, | |
| "step": 74400 | |
| }, | |
| { | |
| "epoch": 21.66779226828132, | |
| "grad_norm": 0.3951086103916168, | |
| "learning_rate": 0.00034018986604542804, | |
| "loss": 3.2172, | |
| "step": 74450 | |
| }, | |
| { | |
| "epoch": 21.682347461574288, | |
| "grad_norm": 0.3802475333213806, | |
| "learning_rate": 0.00034001514269073964, | |
| "loss": 3.2316, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 21.696902654867255, | |
| "grad_norm": 0.3902455270290375, | |
| "learning_rate": 0.0003398404193360512, | |
| "loss": 3.2308, | |
| "step": 74550 | |
| }, | |
| { | |
| "epoch": 21.711457848160222, | |
| "grad_norm": 0.3920290172100067, | |
| "learning_rate": 0.00033966569598136283, | |
| "loss": 3.2337, | |
| "step": 74600 | |
| }, | |
| { | |
| "epoch": 21.72601304145319, | |
| "grad_norm": 0.43754345178604126, | |
| "learning_rate": 0.0003394909726266744, | |
| "loss": 3.2357, | |
| "step": 74650 | |
| }, | |
| { | |
| "epoch": 21.740568234746156, | |
| "grad_norm": 0.38823387026786804, | |
| "learning_rate": 0.000339316249271986, | |
| "loss": 3.2272, | |
| "step": 74700 | |
| }, | |
| { | |
| "epoch": 21.755123428039123, | |
| "grad_norm": 0.41571202874183655, | |
| "learning_rate": 0.00033914152591729756, | |
| "loss": 3.2329, | |
| "step": 74750 | |
| }, | |
| { | |
| "epoch": 21.76967862133209, | |
| "grad_norm": 0.39350152015686035, | |
| "learning_rate": 0.00033896680256260915, | |
| "loss": 3.2326, | |
| "step": 74800 | |
| }, | |
| { | |
| "epoch": 21.784233814625058, | |
| "grad_norm": 0.39219316840171814, | |
| "learning_rate": 0.00033879207920792074, | |
| "loss": 3.2332, | |
| "step": 74850 | |
| }, | |
| { | |
| "epoch": 21.798789007918025, | |
| "grad_norm": 0.4104636013507843, | |
| "learning_rate": 0.0003386173558532324, | |
| "loss": 3.2317, | |
| "step": 74900 | |
| }, | |
| { | |
| "epoch": 21.81334420121099, | |
| "grad_norm": 0.4001675248146057, | |
| "learning_rate": 0.00033844263249854393, | |
| "loss": 3.2335, | |
| "step": 74950 | |
| }, | |
| { | |
| "epoch": 21.82789939450396, | |
| "grad_norm": 0.4079340100288391, | |
| "learning_rate": 0.00033826790914385553, | |
| "loss": 3.2285, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.82789939450396, | |
| "eval_accuracy": 0.374510983573091, | |
| "eval_loss": 3.536165714263916, | |
| "eval_runtime": 82.7532, | |
| "eval_samples_per_second": 201.213, | |
| "eval_steps_per_second": 12.58, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.842454587796926, | |
| "grad_norm": 0.4301753342151642, | |
| "learning_rate": 0.0003380931857891671, | |
| "loss": 3.2349, | |
| "step": 75050 | |
| }, | |
| { | |
| "epoch": 21.857009781089893, | |
| "grad_norm": 0.3775196969509125, | |
| "learning_rate": 0.00033791846243447866, | |
| "loss": 3.2433, | |
| "step": 75100 | |
| }, | |
| { | |
| "epoch": 21.87156497438286, | |
| "grad_norm": 0.44254809617996216, | |
| "learning_rate": 0.0003377437390797903, | |
| "loss": 3.2476, | |
| "step": 75150 | |
| }, | |
| { | |
| "epoch": 21.886120167675827, | |
| "grad_norm": 0.3813041150569916, | |
| "learning_rate": 0.0003375690157251019, | |
| "loss": 3.2233, | |
| "step": 75200 | |
| }, | |
| { | |
| "epoch": 21.900675360968794, | |
| "grad_norm": 0.41615453362464905, | |
| "learning_rate": 0.0003373942923704135, | |
| "loss": 3.2442, | |
| "step": 75250 | |
| }, | |
| { | |
| "epoch": 21.91523055426176, | |
| "grad_norm": 0.39513370394706726, | |
| "learning_rate": 0.00033721956901572504, | |
| "loss": 3.2421, | |
| "step": 75300 | |
| }, | |
| { | |
| "epoch": 21.929785747554728, | |
| "grad_norm": 0.3964758515357971, | |
| "learning_rate": 0.00033704484566103663, | |
| "loss": 3.2363, | |
| "step": 75350 | |
| }, | |
| { | |
| "epoch": 21.944340940847695, | |
| "grad_norm": 0.4087502360343933, | |
| "learning_rate": 0.00033687012230634823, | |
| "loss": 3.2514, | |
| "step": 75400 | |
| }, | |
| { | |
| "epoch": 21.958896134140662, | |
| "grad_norm": 0.41822513937950134, | |
| "learning_rate": 0.0003366953989516599, | |
| "loss": 3.2366, | |
| "step": 75450 | |
| }, | |
| { | |
| "epoch": 21.97345132743363, | |
| "grad_norm": 0.38814598321914673, | |
| "learning_rate": 0.00033652067559697147, | |
| "loss": 3.236, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 21.988006520726596, | |
| "grad_norm": 0.3959311246871948, | |
| "learning_rate": 0.000336345952242283, | |
| "loss": 3.2381, | |
| "step": 75550 | |
| }, | |
| { | |
| "epoch": 22.002328830926874, | |
| "grad_norm": 0.45785027742385864, | |
| "learning_rate": 0.0003361712288875946, | |
| "loss": 3.2214, | |
| "step": 75600 | |
| }, | |
| { | |
| "epoch": 22.01688402421984, | |
| "grad_norm": 0.3880925178527832, | |
| "learning_rate": 0.0003359965055329062, | |
| "loss": 3.1345, | |
| "step": 75650 | |
| }, | |
| { | |
| "epoch": 22.03143921751281, | |
| "grad_norm": 0.432598739862442, | |
| "learning_rate": 0.00033582178217821785, | |
| "loss": 3.1441, | |
| "step": 75700 | |
| }, | |
| { | |
| "epoch": 22.045994410805775, | |
| "grad_norm": 0.3993528485298157, | |
| "learning_rate": 0.0003356470588235294, | |
| "loss": 3.147, | |
| "step": 75750 | |
| }, | |
| { | |
| "epoch": 22.060549604098743, | |
| "grad_norm": 0.3990563750267029, | |
| "learning_rate": 0.000335472335468841, | |
| "loss": 3.1583, | |
| "step": 75800 | |
| }, | |
| { | |
| "epoch": 22.07510479739171, | |
| "grad_norm": 0.42120495438575745, | |
| "learning_rate": 0.0003352976121141526, | |
| "loss": 3.1561, | |
| "step": 75850 | |
| }, | |
| { | |
| "epoch": 22.089659990684677, | |
| "grad_norm": 0.40944042801856995, | |
| "learning_rate": 0.0003351228887594641, | |
| "loss": 3.1592, | |
| "step": 75900 | |
| }, | |
| { | |
| "epoch": 22.104215183977644, | |
| "grad_norm": 0.4035565257072449, | |
| "learning_rate": 0.0003349481654047757, | |
| "loss": 3.159, | |
| "step": 75950 | |
| }, | |
| { | |
| "epoch": 22.11877037727061, | |
| "grad_norm": 0.39391037821769714, | |
| "learning_rate": 0.00033477344205008736, | |
| "loss": 3.1565, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.11877037727061, | |
| "eval_accuracy": 0.37364151656764794, | |
| "eval_loss": 3.550374984741211, | |
| "eval_runtime": 82.6988, | |
| "eval_samples_per_second": 201.345, | |
| "eval_steps_per_second": 12.588, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.133325570563578, | |
| "grad_norm": 0.3885456919670105, | |
| "learning_rate": 0.00033459871869539895, | |
| "loss": 3.1664, | |
| "step": 76050 | |
| }, | |
| { | |
| "epoch": 22.147880763856545, | |
| "grad_norm": 0.39827215671539307, | |
| "learning_rate": 0.0003344239953407105, | |
| "loss": 3.1721, | |
| "step": 76100 | |
| }, | |
| { | |
| "epoch": 22.162435957149512, | |
| "grad_norm": 0.3935714364051819, | |
| "learning_rate": 0.0003342492719860221, | |
| "loss": 3.1665, | |
| "step": 76150 | |
| }, | |
| { | |
| "epoch": 22.17699115044248, | |
| "grad_norm": 0.407362163066864, | |
| "learning_rate": 0.0003340745486313337, | |
| "loss": 3.1765, | |
| "step": 76200 | |
| }, | |
| { | |
| "epoch": 22.191546343735446, | |
| "grad_norm": 0.4094945788383484, | |
| "learning_rate": 0.0003338998252766453, | |
| "loss": 3.1851, | |
| "step": 76250 | |
| }, | |
| { | |
| "epoch": 22.206101537028413, | |
| "grad_norm": 0.46139785647392273, | |
| "learning_rate": 0.00033372510192195687, | |
| "loss": 3.1686, | |
| "step": 76300 | |
| }, | |
| { | |
| "epoch": 22.22065673032138, | |
| "grad_norm": 0.5236421227455139, | |
| "learning_rate": 0.00033355037856726847, | |
| "loss": 3.1716, | |
| "step": 76350 | |
| }, | |
| { | |
| "epoch": 22.235211923614347, | |
| "grad_norm": 0.41476672887802124, | |
| "learning_rate": 0.00033337565521258006, | |
| "loss": 3.1791, | |
| "step": 76400 | |
| }, | |
| { | |
| "epoch": 22.24976711690731, | |
| "grad_norm": 0.43234819173812866, | |
| "learning_rate": 0.00033320093185789165, | |
| "loss": 3.1801, | |
| "step": 76450 | |
| }, | |
| { | |
| "epoch": 22.264322310200278, | |
| "grad_norm": 0.3955042064189911, | |
| "learning_rate": 0.0003330262085032032, | |
| "loss": 3.1834, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 22.278877503493245, | |
| "grad_norm": 0.4020642638206482, | |
| "learning_rate": 0.00033285148514851484, | |
| "loss": 3.1818, | |
| "step": 76550 | |
| }, | |
| { | |
| "epoch": 22.293432696786212, | |
| "grad_norm": 0.395889014005661, | |
| "learning_rate": 0.00033267676179382644, | |
| "loss": 3.1904, | |
| "step": 76600 | |
| }, | |
| { | |
| "epoch": 22.30798789007918, | |
| "grad_norm": 0.45003068447113037, | |
| "learning_rate": 0.00033250203843913803, | |
| "loss": 3.1897, | |
| "step": 76650 | |
| }, | |
| { | |
| "epoch": 22.322543083372146, | |
| "grad_norm": 0.4547971487045288, | |
| "learning_rate": 0.00033232731508444957, | |
| "loss": 3.1954, | |
| "step": 76700 | |
| }, | |
| { | |
| "epoch": 22.337098276665113, | |
| "grad_norm": 0.41432705521583557, | |
| "learning_rate": 0.00033215259172976117, | |
| "loss": 3.1905, | |
| "step": 76750 | |
| }, | |
| { | |
| "epoch": 22.35165346995808, | |
| "grad_norm": 0.3937246799468994, | |
| "learning_rate": 0.00033197786837507276, | |
| "loss": 3.1905, | |
| "step": 76800 | |
| }, | |
| { | |
| "epoch": 22.366208663251047, | |
| "grad_norm": 0.42874905467033386, | |
| "learning_rate": 0.0003318031450203844, | |
| "loss": 3.1962, | |
| "step": 76850 | |
| }, | |
| { | |
| "epoch": 22.380763856544014, | |
| "grad_norm": 0.43876227736473083, | |
| "learning_rate": 0.00033162842166569595, | |
| "loss": 3.1871, | |
| "step": 76900 | |
| }, | |
| { | |
| "epoch": 22.39531904983698, | |
| "grad_norm": 0.40947768092155457, | |
| "learning_rate": 0.00033145369831100754, | |
| "loss": 3.1985, | |
| "step": 76950 | |
| }, | |
| { | |
| "epoch": 22.40987424312995, | |
| "grad_norm": 0.4180110692977905, | |
| "learning_rate": 0.00033127897495631914, | |
| "loss": 3.1989, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.40987424312995, | |
| "eval_accuracy": 0.37402500816520956, | |
| "eval_loss": 3.5429611206054688, | |
| "eval_runtime": 82.8421, | |
| "eval_samples_per_second": 200.997, | |
| "eval_steps_per_second": 12.566, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.424429436422916, | |
| "grad_norm": 0.40657177567481995, | |
| "learning_rate": 0.0003311042516016307, | |
| "loss": 3.2016, | |
| "step": 77050 | |
| }, | |
| { | |
| "epoch": 22.438984629715883, | |
| "grad_norm": 0.4393850862979889, | |
| "learning_rate": 0.0003309295282469423, | |
| "loss": 3.1983, | |
| "step": 77100 | |
| }, | |
| { | |
| "epoch": 22.45353982300885, | |
| "grad_norm": 0.4241443872451782, | |
| "learning_rate": 0.0003307548048922539, | |
| "loss": 3.2001, | |
| "step": 77150 | |
| }, | |
| { | |
| "epoch": 22.468095016301817, | |
| "grad_norm": 0.40221163630485535, | |
| "learning_rate": 0.0003305800815375655, | |
| "loss": 3.2039, | |
| "step": 77200 | |
| }, | |
| { | |
| "epoch": 22.482650209594784, | |
| "grad_norm": 0.4283682703971863, | |
| "learning_rate": 0.00033040535818287705, | |
| "loss": 3.2032, | |
| "step": 77250 | |
| }, | |
| { | |
| "epoch": 22.49720540288775, | |
| "grad_norm": 0.44070905447006226, | |
| "learning_rate": 0.00033023063482818865, | |
| "loss": 3.2019, | |
| "step": 77300 | |
| }, | |
| { | |
| "epoch": 22.511760596180718, | |
| "grad_norm": 0.4331526756286621, | |
| "learning_rate": 0.00033005591147350024, | |
| "loss": 3.2005, | |
| "step": 77350 | |
| }, | |
| { | |
| "epoch": 22.526315789473685, | |
| "grad_norm": 0.4136602580547333, | |
| "learning_rate": 0.0003298811881188119, | |
| "loss": 3.2147, | |
| "step": 77400 | |
| }, | |
| { | |
| "epoch": 22.540870982766652, | |
| "grad_norm": 0.42045098543167114, | |
| "learning_rate": 0.00032970646476412343, | |
| "loss": 3.195, | |
| "step": 77450 | |
| }, | |
| { | |
| "epoch": 22.55542617605962, | |
| "grad_norm": 0.39976319670677185, | |
| "learning_rate": 0.000329531741409435, | |
| "loss": 3.2131, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 22.569981369352586, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0003293570180547466, | |
| "loss": 3.2018, | |
| "step": 77550 | |
| }, | |
| { | |
| "epoch": 22.584536562645553, | |
| "grad_norm": 0.4044171869754791, | |
| "learning_rate": 0.0003291822947000582, | |
| "loss": 3.2077, | |
| "step": 77600 | |
| }, | |
| { | |
| "epoch": 22.59909175593852, | |
| "grad_norm": 0.4308306872844696, | |
| "learning_rate": 0.00032900757134536975, | |
| "loss": 3.2122, | |
| "step": 77650 | |
| }, | |
| { | |
| "epoch": 22.613646949231487, | |
| "grad_norm": 0.4276672601699829, | |
| "learning_rate": 0.0003288328479906814, | |
| "loss": 3.2164, | |
| "step": 77700 | |
| }, | |
| { | |
| "epoch": 22.628202142524454, | |
| "grad_norm": 0.41239678859710693, | |
| "learning_rate": 0.000328658124635993, | |
| "loss": 3.2153, | |
| "step": 77750 | |
| }, | |
| { | |
| "epoch": 22.64275733581742, | |
| "grad_norm": 0.3834725618362427, | |
| "learning_rate": 0.0003284834012813046, | |
| "loss": 3.2053, | |
| "step": 77800 | |
| }, | |
| { | |
| "epoch": 22.657312529110385, | |
| "grad_norm": 0.4435277581214905, | |
| "learning_rate": 0.00032830867792661613, | |
| "loss": 3.2152, | |
| "step": 77850 | |
| }, | |
| { | |
| "epoch": 22.671867722403352, | |
| "grad_norm": 0.4278111457824707, | |
| "learning_rate": 0.0003281339545719277, | |
| "loss": 3.2238, | |
| "step": 77900 | |
| }, | |
| { | |
| "epoch": 22.68642291569632, | |
| "grad_norm": 0.4114883542060852, | |
| "learning_rate": 0.0003279592312172394, | |
| "loss": 3.216, | |
| "step": 77950 | |
| }, | |
| { | |
| "epoch": 22.700978108989286, | |
| "grad_norm": 0.38967975974082947, | |
| "learning_rate": 0.00032778450786255097, | |
| "loss": 3.2211, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.700978108989286, | |
| "eval_accuracy": 0.37451615477452915, | |
| "eval_loss": 3.5369396209716797, | |
| "eval_runtime": 82.7867, | |
| "eval_samples_per_second": 201.131, | |
| "eval_steps_per_second": 12.574, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.715533302282253, | |
| "grad_norm": 0.3989904522895813, | |
| "learning_rate": 0.0003276097845078625, | |
| "loss": 3.2045, | |
| "step": 78050 | |
| }, | |
| { | |
| "epoch": 22.73008849557522, | |
| "grad_norm": 0.44048959016799927, | |
| "learning_rate": 0.0003274350611531741, | |
| "loss": 3.2292, | |
| "step": 78100 | |
| }, | |
| { | |
| "epoch": 22.744643688868187, | |
| "grad_norm": 0.40299227833747864, | |
| "learning_rate": 0.0003272603377984857, | |
| "loss": 3.2197, | |
| "step": 78150 | |
| }, | |
| { | |
| "epoch": 22.759198882161154, | |
| "grad_norm": 0.3916628658771515, | |
| "learning_rate": 0.00032708561444379724, | |
| "loss": 3.2165, | |
| "step": 78200 | |
| }, | |
| { | |
| "epoch": 22.77375407545412, | |
| "grad_norm": 0.4167213439941406, | |
| "learning_rate": 0.0003269108910891089, | |
| "loss": 3.2249, | |
| "step": 78250 | |
| }, | |
| { | |
| "epoch": 22.78830926874709, | |
| "grad_norm": 0.40860942006111145, | |
| "learning_rate": 0.0003267361677344205, | |
| "loss": 3.2213, | |
| "step": 78300 | |
| }, | |
| { | |
| "epoch": 22.802864462040056, | |
| "grad_norm": 0.40919870138168335, | |
| "learning_rate": 0.0003265614443797321, | |
| "loss": 3.2269, | |
| "step": 78350 | |
| }, | |
| { | |
| "epoch": 22.817419655333023, | |
| "grad_norm": 0.4579060971736908, | |
| "learning_rate": 0.0003263867210250436, | |
| "loss": 3.2277, | |
| "step": 78400 | |
| }, | |
| { | |
| "epoch": 22.83197484862599, | |
| "grad_norm": 0.4217158257961273, | |
| "learning_rate": 0.0003262119976703552, | |
| "loss": 3.2233, | |
| "step": 78450 | |
| }, | |
| { | |
| "epoch": 22.846530041918957, | |
| "grad_norm": 0.37889739871025085, | |
| "learning_rate": 0.00032603727431566686, | |
| "loss": 3.2297, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 22.861085235211924, | |
| "grad_norm": 0.4110410809516907, | |
| "learning_rate": 0.00032586255096097845, | |
| "loss": 3.2405, | |
| "step": 78550 | |
| }, | |
| { | |
| "epoch": 22.87564042850489, | |
| "grad_norm": 0.40812453627586365, | |
| "learning_rate": 0.00032568782760629005, | |
| "loss": 3.2298, | |
| "step": 78600 | |
| }, | |
| { | |
| "epoch": 22.890195621797858, | |
| "grad_norm": 0.42722323536872864, | |
| "learning_rate": 0.0003255131042516016, | |
| "loss": 3.2383, | |
| "step": 78650 | |
| }, | |
| { | |
| "epoch": 22.904750815090825, | |
| "grad_norm": 0.43514180183410645, | |
| "learning_rate": 0.0003253383808969132, | |
| "loss": 3.2225, | |
| "step": 78700 | |
| }, | |
| { | |
| "epoch": 22.919306008383792, | |
| "grad_norm": 0.4373376667499542, | |
| "learning_rate": 0.0003251636575422248, | |
| "loss": 3.2353, | |
| "step": 78750 | |
| }, | |
| { | |
| "epoch": 22.93386120167676, | |
| "grad_norm": 0.4121592938899994, | |
| "learning_rate": 0.0003249889341875364, | |
| "loss": 3.2231, | |
| "step": 78800 | |
| }, | |
| { | |
| "epoch": 22.948416394969726, | |
| "grad_norm": 0.38480544090270996, | |
| "learning_rate": 0.00032481421083284796, | |
| "loss": 3.2334, | |
| "step": 78850 | |
| }, | |
| { | |
| "epoch": 22.962971588262693, | |
| "grad_norm": 0.41385582089424133, | |
| "learning_rate": 0.00032463948747815956, | |
| "loss": 3.2443, | |
| "step": 78900 | |
| }, | |
| { | |
| "epoch": 22.97752678155566, | |
| "grad_norm": 0.4353277385234833, | |
| "learning_rate": 0.00032446476412347115, | |
| "loss": 3.2282, | |
| "step": 78950 | |
| }, | |
| { | |
| "epoch": 22.992081974848627, | |
| "grad_norm": 0.415485680103302, | |
| "learning_rate": 0.0003242900407687827, | |
| "loss": 3.2442, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 22.992081974848627, | |
| "eval_accuracy": 0.3749589976613241, | |
| "eval_loss": 3.528618335723877, | |
| "eval_runtime": 82.9929, | |
| "eval_samples_per_second": 200.632, | |
| "eval_steps_per_second": 12.543, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 23.006404285048905, | |
| "grad_norm": 0.4230590760707855, | |
| "learning_rate": 0.00032411531741409434, | |
| "loss": 3.1887, | |
| "step": 79050 | |
| }, | |
| { | |
| "epoch": 23.020959478341872, | |
| "grad_norm": 0.4250546395778656, | |
| "learning_rate": 0.00032394059405940593, | |
| "loss": 3.1257, | |
| "step": 79100 | |
| }, | |
| { | |
| "epoch": 23.03551467163484, | |
| "grad_norm": 0.4422360360622406, | |
| "learning_rate": 0.00032376587070471753, | |
| "loss": 3.1264, | |
| "step": 79150 | |
| }, | |
| { | |
| "epoch": 23.050069864927806, | |
| "grad_norm": 0.45506057143211365, | |
| "learning_rate": 0.00032359114735002907, | |
| "loss": 3.1343, | |
| "step": 79200 | |
| }, | |
| { | |
| "epoch": 23.064625058220773, | |
| "grad_norm": 0.40860608220100403, | |
| "learning_rate": 0.00032341642399534066, | |
| "loss": 3.1479, | |
| "step": 79250 | |
| }, | |
| { | |
| "epoch": 23.07918025151374, | |
| "grad_norm": 0.4227176010608673, | |
| "learning_rate": 0.00032324170064065226, | |
| "loss": 3.1292, | |
| "step": 79300 | |
| }, | |
| { | |
| "epoch": 23.093735444806708, | |
| "grad_norm": 0.44888702034950256, | |
| "learning_rate": 0.0003230669772859639, | |
| "loss": 3.1605, | |
| "step": 79350 | |
| }, | |
| { | |
| "epoch": 23.108290638099675, | |
| "grad_norm": 0.3960554301738739, | |
| "learning_rate": 0.00032289225393127545, | |
| "loss": 3.1467, | |
| "step": 79400 | |
| }, | |
| { | |
| "epoch": 23.12284583139264, | |
| "grad_norm": 0.42449507117271423, | |
| "learning_rate": 0.00032271753057658704, | |
| "loss": 3.148, | |
| "step": 79450 | |
| }, | |
| { | |
| "epoch": 23.13740102468561, | |
| "grad_norm": 0.41957584023475647, | |
| "learning_rate": 0.00032254280722189863, | |
| "loss": 3.162, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 23.151956217978576, | |
| "grad_norm": 0.45502346754074097, | |
| "learning_rate": 0.00032236808386721023, | |
| "loss": 3.1498, | |
| "step": 79550 | |
| }, | |
| { | |
| "epoch": 23.166511411271543, | |
| "grad_norm": 0.41477227210998535, | |
| "learning_rate": 0.00032219336051252177, | |
| "loss": 3.1524, | |
| "step": 79600 | |
| }, | |
| { | |
| "epoch": 23.18106660456451, | |
| "grad_norm": 0.40256333351135254, | |
| "learning_rate": 0.0003220186371578334, | |
| "loss": 3.1561, | |
| "step": 79650 | |
| }, | |
| { | |
| "epoch": 23.195621797857477, | |
| "grad_norm": 0.4273405969142914, | |
| "learning_rate": 0.000321843913803145, | |
| "loss": 3.1723, | |
| "step": 79700 | |
| }, | |
| { | |
| "epoch": 23.210176991150444, | |
| "grad_norm": 0.4353051781654358, | |
| "learning_rate": 0.0003216691904484566, | |
| "loss": 3.1656, | |
| "step": 79750 | |
| }, | |
| { | |
| "epoch": 23.22473218444341, | |
| "grad_norm": 0.4311169385910034, | |
| "learning_rate": 0.00032149446709376815, | |
| "loss": 3.1724, | |
| "step": 79800 | |
| }, | |
| { | |
| "epoch": 23.239287377736375, | |
| "grad_norm": 0.4456189274787903, | |
| "learning_rate": 0.00032131974373907974, | |
| "loss": 3.1747, | |
| "step": 79850 | |
| }, | |
| { | |
| "epoch": 23.25384257102934, | |
| "grad_norm": 0.41731199622154236, | |
| "learning_rate": 0.0003211450203843914, | |
| "loss": 3.1804, | |
| "step": 79900 | |
| }, | |
| { | |
| "epoch": 23.26839776432231, | |
| "grad_norm": 0.4556697607040405, | |
| "learning_rate": 0.000320970297029703, | |
| "loss": 3.1739, | |
| "step": 79950 | |
| }, | |
| { | |
| "epoch": 23.282952957615276, | |
| "grad_norm": 0.42701220512390137, | |
| "learning_rate": 0.0003207955736750145, | |
| "loss": 3.1825, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.282952957615276, | |
| "eval_accuracy": 0.373897843620753, | |
| "eval_loss": 3.5500452518463135, | |
| "eval_runtime": 83.2551, | |
| "eval_samples_per_second": 200.0, | |
| "eval_steps_per_second": 12.504, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.297508150908243, | |
| "grad_norm": 0.435136079788208, | |
| "learning_rate": 0.0003206208503203261, | |
| "loss": 3.1813, | |
| "step": 80050 | |
| }, | |
| { | |
| "epoch": 23.31206334420121, | |
| "grad_norm": 0.3913682997226715, | |
| "learning_rate": 0.0003204461269656377, | |
| "loss": 3.1752, | |
| "step": 80100 | |
| }, | |
| { | |
| "epoch": 23.326618537494177, | |
| "grad_norm": 0.4277016222476959, | |
| "learning_rate": 0.00032027140361094925, | |
| "loss": 3.1926, | |
| "step": 80150 | |
| }, | |
| { | |
| "epoch": 23.341173730787144, | |
| "grad_norm": 0.40360701084136963, | |
| "learning_rate": 0.0003200966802562609, | |
| "loss": 3.1885, | |
| "step": 80200 | |
| }, | |
| { | |
| "epoch": 23.35572892408011, | |
| "grad_norm": 0.4251612424850464, | |
| "learning_rate": 0.0003199219569015725, | |
| "loss": 3.187, | |
| "step": 80250 | |
| }, | |
| { | |
| "epoch": 23.370284117373078, | |
| "grad_norm": 0.42847540974617004, | |
| "learning_rate": 0.0003197472335468841, | |
| "loss": 3.1803, | |
| "step": 80300 | |
| }, | |
| { | |
| "epoch": 23.384839310666045, | |
| "grad_norm": 0.4316292107105255, | |
| "learning_rate": 0.00031957251019219563, | |
| "loss": 3.1869, | |
| "step": 80350 | |
| }, | |
| { | |
| "epoch": 23.399394503959012, | |
| "grad_norm": 0.4152030944824219, | |
| "learning_rate": 0.0003193977868375072, | |
| "loss": 3.1874, | |
| "step": 80400 | |
| }, | |
| { | |
| "epoch": 23.41394969725198, | |
| "grad_norm": 0.40146902203559875, | |
| "learning_rate": 0.00031922306348281887, | |
| "loss": 3.1849, | |
| "step": 80450 | |
| }, | |
| { | |
| "epoch": 23.428504890544946, | |
| "grad_norm": 0.41296109557151794, | |
| "learning_rate": 0.00031904834012813047, | |
| "loss": 3.1837, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 23.443060083837914, | |
| "grad_norm": 0.4590873718261719, | |
| "learning_rate": 0.000318873616773442, | |
| "loss": 3.201, | |
| "step": 80550 | |
| }, | |
| { | |
| "epoch": 23.45761527713088, | |
| "grad_norm": 0.40355953574180603, | |
| "learning_rate": 0.0003186988934187536, | |
| "loss": 3.1957, | |
| "step": 80600 | |
| }, | |
| { | |
| "epoch": 23.472170470423848, | |
| "grad_norm": 0.4102541506290436, | |
| "learning_rate": 0.0003185241700640652, | |
| "loss": 3.1961, | |
| "step": 80650 | |
| }, | |
| { | |
| "epoch": 23.486725663716815, | |
| "grad_norm": 0.4047577381134033, | |
| "learning_rate": 0.0003183494467093768, | |
| "loss": 3.2013, | |
| "step": 80700 | |
| }, | |
| { | |
| "epoch": 23.50128085700978, | |
| "grad_norm": 0.43577125668525696, | |
| "learning_rate": 0.0003181747233546884, | |
| "loss": 3.1996, | |
| "step": 80750 | |
| }, | |
| { | |
| "epoch": 23.51583605030275, | |
| "grad_norm": 0.41916581988334656, | |
| "learning_rate": 0.000318, | |
| "loss": 3.2011, | |
| "step": 80800 | |
| }, | |
| { | |
| "epoch": 23.530391243595716, | |
| "grad_norm": 0.38767364621162415, | |
| "learning_rate": 0.00031782527664531157, | |
| "loss": 3.1961, | |
| "step": 80850 | |
| }, | |
| { | |
| "epoch": 23.544946436888683, | |
| "grad_norm": 0.42400556802749634, | |
| "learning_rate": 0.00031765055329062317, | |
| "loss": 3.2012, | |
| "step": 80900 | |
| }, | |
| { | |
| "epoch": 23.55950163018165, | |
| "grad_norm": 0.40898025035858154, | |
| "learning_rate": 0.0003174758299359347, | |
| "loss": 3.2032, | |
| "step": 80950 | |
| }, | |
| { | |
| "epoch": 23.574056823474617, | |
| "grad_norm": 0.4222612977027893, | |
| "learning_rate": 0.0003173011065812463, | |
| "loss": 3.1975, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.574056823474617, | |
| "eval_accuracy": 0.3741784988260785, | |
| "eval_loss": 3.5403919219970703, | |
| "eval_runtime": 82.7958, | |
| "eval_samples_per_second": 201.109, | |
| "eval_steps_per_second": 12.573, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.588612016767584, | |
| "grad_norm": 0.4143919050693512, | |
| "learning_rate": 0.00031712638322655795, | |
| "loss": 3.1867, | |
| "step": 81050 | |
| }, | |
| { | |
| "epoch": 23.60316721006055, | |
| "grad_norm": 0.41058212518692017, | |
| "learning_rate": 0.00031695165987186954, | |
| "loss": 3.1988, | |
| "step": 81100 | |
| }, | |
| { | |
| "epoch": 23.61772240335352, | |
| "grad_norm": 0.42099881172180176, | |
| "learning_rate": 0.0003167769365171811, | |
| "loss": 3.201, | |
| "step": 81150 | |
| }, | |
| { | |
| "epoch": 23.63227759664648, | |
| "grad_norm": 0.41187793016433716, | |
| "learning_rate": 0.0003166022131624927, | |
| "loss": 3.2033, | |
| "step": 81200 | |
| }, | |
| { | |
| "epoch": 23.64683278993945, | |
| "grad_norm": 0.40702155232429504, | |
| "learning_rate": 0.00031642748980780427, | |
| "loss": 3.2016, | |
| "step": 81250 | |
| }, | |
| { | |
| "epoch": 23.661387983232416, | |
| "grad_norm": 0.42631128430366516, | |
| "learning_rate": 0.0003162527664531159, | |
| "loss": 3.2047, | |
| "step": 81300 | |
| }, | |
| { | |
| "epoch": 23.675943176525383, | |
| "grad_norm": 0.5031477212905884, | |
| "learning_rate": 0.00031607804309842746, | |
| "loss": 3.1968, | |
| "step": 81350 | |
| }, | |
| { | |
| "epoch": 23.69049836981835, | |
| "grad_norm": 0.41763603687286377, | |
| "learning_rate": 0.00031590331974373905, | |
| "loss": 3.2086, | |
| "step": 81400 | |
| }, | |
| { | |
| "epoch": 23.705053563111317, | |
| "grad_norm": 0.44476330280303955, | |
| "learning_rate": 0.00031572859638905065, | |
| "loss": 3.2037, | |
| "step": 81450 | |
| }, | |
| { | |
| "epoch": 23.719608756404284, | |
| "grad_norm": 0.3911072313785553, | |
| "learning_rate": 0.0003155538730343622, | |
| "loss": 3.2128, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 23.73416394969725, | |
| "grad_norm": 0.4476863145828247, | |
| "learning_rate": 0.0003153791496796738, | |
| "loss": 3.2224, | |
| "step": 81550 | |
| }, | |
| { | |
| "epoch": 23.74871914299022, | |
| "grad_norm": 0.4151424169540405, | |
| "learning_rate": 0.00031520442632498543, | |
| "loss": 3.2139, | |
| "step": 81600 | |
| }, | |
| { | |
| "epoch": 23.763274336283185, | |
| "grad_norm": 0.4364260733127594, | |
| "learning_rate": 0.000315029702970297, | |
| "loss": 3.2067, | |
| "step": 81650 | |
| }, | |
| { | |
| "epoch": 23.777829529576152, | |
| "grad_norm": 0.3949473798274994, | |
| "learning_rate": 0.0003148549796156086, | |
| "loss": 3.2081, | |
| "step": 81700 | |
| }, | |
| { | |
| "epoch": 23.79238472286912, | |
| "grad_norm": 0.4132239520549774, | |
| "learning_rate": 0.00031468025626092016, | |
| "loss": 3.2105, | |
| "step": 81750 | |
| }, | |
| { | |
| "epoch": 23.806939916162086, | |
| "grad_norm": 0.4394204318523407, | |
| "learning_rate": 0.00031450553290623175, | |
| "loss": 3.2174, | |
| "step": 81800 | |
| }, | |
| { | |
| "epoch": 23.821495109455054, | |
| "grad_norm": 0.43531525135040283, | |
| "learning_rate": 0.0003143308095515434, | |
| "loss": 3.2154, | |
| "step": 81850 | |
| }, | |
| { | |
| "epoch": 23.83605030274802, | |
| "grad_norm": 0.4317162036895752, | |
| "learning_rate": 0.000314156086196855, | |
| "loss": 3.2183, | |
| "step": 81900 | |
| }, | |
| { | |
| "epoch": 23.850605496040988, | |
| "grad_norm": 0.44531089067459106, | |
| "learning_rate": 0.00031398136284216654, | |
| "loss": 3.2179, | |
| "step": 81950 | |
| }, | |
| { | |
| "epoch": 23.865160689333955, | |
| "grad_norm": 0.40127038955688477, | |
| "learning_rate": 0.00031380663948747813, | |
| "loss": 3.227, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.865160689333955, | |
| "eval_accuracy": 0.37466153605132463, | |
| "eval_loss": 3.5329294204711914, | |
| "eval_runtime": 82.7648, | |
| "eval_samples_per_second": 201.185, | |
| "eval_steps_per_second": 12.578, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.879715882626922, | |
| "grad_norm": 0.40856117010116577, | |
| "learning_rate": 0.0003136319161327897, | |
| "loss": 3.2189, | |
| "step": 82050 | |
| }, | |
| { | |
| "epoch": 23.89427107591989, | |
| "grad_norm": 0.3996933698654175, | |
| "learning_rate": 0.00031345719277810127, | |
| "loss": 3.2319, | |
| "step": 82100 | |
| }, | |
| { | |
| "epoch": 23.908826269212856, | |
| "grad_norm": 0.42883631587028503, | |
| "learning_rate": 0.0003132824694234129, | |
| "loss": 3.2266, | |
| "step": 82150 | |
| }, | |
| { | |
| "epoch": 23.923381462505823, | |
| "grad_norm": 0.4053521454334259, | |
| "learning_rate": 0.0003131077460687245, | |
| "loss": 3.2229, | |
| "step": 82200 | |
| }, | |
| { | |
| "epoch": 23.93793665579879, | |
| "grad_norm": 0.42732861638069153, | |
| "learning_rate": 0.0003129330227140361, | |
| "loss": 3.2188, | |
| "step": 82250 | |
| }, | |
| { | |
| "epoch": 23.952491849091757, | |
| "grad_norm": 0.4492589831352234, | |
| "learning_rate": 0.00031275829935934764, | |
| "loss": 3.2222, | |
| "step": 82300 | |
| }, | |
| { | |
| "epoch": 23.967047042384724, | |
| "grad_norm": 0.42206841707229614, | |
| "learning_rate": 0.00031258357600465924, | |
| "loss": 3.2176, | |
| "step": 82350 | |
| }, | |
| { | |
| "epoch": 23.98160223567769, | |
| "grad_norm": 0.45304062962532043, | |
| "learning_rate": 0.00031240885264997083, | |
| "loss": 3.2368, | |
| "step": 82400 | |
| }, | |
| { | |
| "epoch": 23.99615742897066, | |
| "grad_norm": 0.4101537764072418, | |
| "learning_rate": 0.0003122341292952825, | |
| "loss": 3.2192, | |
| "step": 82450 | |
| }, | |
| { | |
| "epoch": 24.010479739170936, | |
| "grad_norm": 0.41270461678504944, | |
| "learning_rate": 0.000312059405940594, | |
| "loss": 3.1534, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 24.025034932463903, | |
| "grad_norm": 0.42348232865333557, | |
| "learning_rate": 0.0003118846825859056, | |
| "loss": 3.1261, | |
| "step": 82550 | |
| }, | |
| { | |
| "epoch": 24.03959012575687, | |
| "grad_norm": 0.4316028356552124, | |
| "learning_rate": 0.0003117099592312172, | |
| "loss": 3.1285, | |
| "step": 82600 | |
| }, | |
| { | |
| "epoch": 24.054145319049837, | |
| "grad_norm": 0.42739230394363403, | |
| "learning_rate": 0.0003115352358765288, | |
| "loss": 3.1259, | |
| "step": 82650 | |
| }, | |
| { | |
| "epoch": 24.068700512342804, | |
| "grad_norm": 0.4323435127735138, | |
| "learning_rate": 0.0003113605125218404, | |
| "loss": 3.1418, | |
| "step": 82700 | |
| }, | |
| { | |
| "epoch": 24.08325570563577, | |
| "grad_norm": 0.4205418527126312, | |
| "learning_rate": 0.000311185789167152, | |
| "loss": 3.1416, | |
| "step": 82750 | |
| }, | |
| { | |
| "epoch": 24.09781089892874, | |
| "grad_norm": 0.42595234513282776, | |
| "learning_rate": 0.0003110110658124636, | |
| "loss": 3.1448, | |
| "step": 82800 | |
| }, | |
| { | |
| "epoch": 24.112366092221706, | |
| "grad_norm": 0.4037402272224426, | |
| "learning_rate": 0.0003108363424577752, | |
| "loss": 3.1493, | |
| "step": 82850 | |
| }, | |
| { | |
| "epoch": 24.126921285514673, | |
| "grad_norm": 0.43444186449050903, | |
| "learning_rate": 0.0003106616191030867, | |
| "loss": 3.143, | |
| "step": 82900 | |
| }, | |
| { | |
| "epoch": 24.14147647880764, | |
| "grad_norm": 0.41186627745628357, | |
| "learning_rate": 0.0003104868957483983, | |
| "loss": 3.1403, | |
| "step": 82950 | |
| }, | |
| { | |
| "epoch": 24.156031672100607, | |
| "grad_norm": 0.3922557532787323, | |
| "learning_rate": 0.00031031217239370996, | |
| "loss": 3.1517, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.156031672100607, | |
| "eval_accuracy": 0.37417802871685685, | |
| "eval_loss": 3.548198938369751, | |
| "eval_runtime": 82.7044, | |
| "eval_samples_per_second": 201.332, | |
| "eval_steps_per_second": 12.587, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.170586865393574, | |
| "grad_norm": 0.4327329099178314, | |
| "learning_rate": 0.00031013744903902156, | |
| "loss": 3.1534, | |
| "step": 83050 | |
| }, | |
| { | |
| "epoch": 24.18514205868654, | |
| "grad_norm": 0.42072615027427673, | |
| "learning_rate": 0.0003099627256843331, | |
| "loss": 3.1604, | |
| "step": 83100 | |
| }, | |
| { | |
| "epoch": 24.199697251979508, | |
| "grad_norm": 0.41045641899108887, | |
| "learning_rate": 0.0003097880023296447, | |
| "loss": 3.1498, | |
| "step": 83150 | |
| }, | |
| { | |
| "epoch": 24.21425244527247, | |
| "grad_norm": 0.4308350086212158, | |
| "learning_rate": 0.0003096132789749563, | |
| "loss": 3.1525, | |
| "step": 83200 | |
| }, | |
| { | |
| "epoch": 24.22880763856544, | |
| "grad_norm": 0.42125487327575684, | |
| "learning_rate": 0.00030943855562026794, | |
| "loss": 3.1504, | |
| "step": 83250 | |
| }, | |
| { | |
| "epoch": 24.243362831858406, | |
| "grad_norm": 0.4198237657546997, | |
| "learning_rate": 0.0003092638322655795, | |
| "loss": 3.1647, | |
| "step": 83300 | |
| }, | |
| { | |
| "epoch": 24.257918025151373, | |
| "grad_norm": 0.4200682044029236, | |
| "learning_rate": 0.00030908910891089107, | |
| "loss": 3.1625, | |
| "step": 83350 | |
| }, | |
| { | |
| "epoch": 24.27247321844434, | |
| "grad_norm": 0.42110204696655273, | |
| "learning_rate": 0.00030891438555620266, | |
| "loss": 3.162, | |
| "step": 83400 | |
| }, | |
| { | |
| "epoch": 24.287028411737307, | |
| "grad_norm": 0.4299944043159485, | |
| "learning_rate": 0.0003087396622015142, | |
| "loss": 3.1751, | |
| "step": 83450 | |
| }, | |
| { | |
| "epoch": 24.301583605030274, | |
| "grad_norm": 0.4246542155742645, | |
| "learning_rate": 0.0003085649388468258, | |
| "loss": 3.1686, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 24.31613879832324, | |
| "grad_norm": 0.4109070897102356, | |
| "learning_rate": 0.00030839021549213745, | |
| "loss": 3.1718, | |
| "step": 83550 | |
| }, | |
| { | |
| "epoch": 24.330693991616208, | |
| "grad_norm": 0.4052143096923828, | |
| "learning_rate": 0.00030821549213744904, | |
| "loss": 3.168, | |
| "step": 83600 | |
| }, | |
| { | |
| "epoch": 24.345249184909175, | |
| "grad_norm": 0.4216053783893585, | |
| "learning_rate": 0.0003080407687827606, | |
| "loss": 3.1781, | |
| "step": 83650 | |
| }, | |
| { | |
| "epoch": 24.359804378202142, | |
| "grad_norm": 0.4117390811443329, | |
| "learning_rate": 0.0003078660454280722, | |
| "loss": 3.1798, | |
| "step": 83700 | |
| }, | |
| { | |
| "epoch": 24.37435957149511, | |
| "grad_norm": 0.4203890562057495, | |
| "learning_rate": 0.00030769132207338377, | |
| "loss": 3.1784, | |
| "step": 83750 | |
| }, | |
| { | |
| "epoch": 24.388914764788076, | |
| "grad_norm": 0.43701109290122986, | |
| "learning_rate": 0.00030751659871869536, | |
| "loss": 3.1883, | |
| "step": 83800 | |
| }, | |
| { | |
| "epoch": 24.403469958081043, | |
| "grad_norm": 0.4347490668296814, | |
| "learning_rate": 0.00030734187536400696, | |
| "loss": 3.1734, | |
| "step": 83850 | |
| }, | |
| { | |
| "epoch": 24.41802515137401, | |
| "grad_norm": 0.4421546459197998, | |
| "learning_rate": 0.00030716715200931855, | |
| "loss": 3.1709, | |
| "step": 83900 | |
| }, | |
| { | |
| "epoch": 24.432580344666977, | |
| "grad_norm": 0.43777987360954285, | |
| "learning_rate": 0.00030699242865463015, | |
| "loss": 3.19, | |
| "step": 83950 | |
| }, | |
| { | |
| "epoch": 24.447135537959944, | |
| "grad_norm": 0.4298378825187683, | |
| "learning_rate": 0.00030681770529994174, | |
| "loss": 3.1805, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.447135537959944, | |
| "eval_accuracy": 0.37418343497290585, | |
| "eval_loss": 3.544612169265747, | |
| "eval_runtime": 82.8561, | |
| "eval_samples_per_second": 200.963, | |
| "eval_steps_per_second": 12.564, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.46169073125291, | |
| "grad_norm": 0.4264633059501648, | |
| "learning_rate": 0.0003066429819452533, | |
| "loss": 3.1822, | |
| "step": 84050 | |
| }, | |
| { | |
| "epoch": 24.47624592454588, | |
| "grad_norm": 0.39448273181915283, | |
| "learning_rate": 0.00030646825859056493, | |
| "loss": 3.1748, | |
| "step": 84100 | |
| }, | |
| { | |
| "epoch": 24.490801117838846, | |
| "grad_norm": 0.4567188024520874, | |
| "learning_rate": 0.0003062935352358765, | |
| "loss": 3.1865, | |
| "step": 84150 | |
| }, | |
| { | |
| "epoch": 24.505356311131813, | |
| "grad_norm": 0.45742517709732056, | |
| "learning_rate": 0.0003061188118811881, | |
| "loss": 3.1852, | |
| "step": 84200 | |
| }, | |
| { | |
| "epoch": 24.51991150442478, | |
| "grad_norm": 0.4224489629268646, | |
| "learning_rate": 0.00030594408852649966, | |
| "loss": 3.1893, | |
| "step": 84250 | |
| }, | |
| { | |
| "epoch": 24.534466697717747, | |
| "grad_norm": 0.3984431326389313, | |
| "learning_rate": 0.00030576936517181125, | |
| "loss": 3.1897, | |
| "step": 84300 | |
| }, | |
| { | |
| "epoch": 24.549021891010714, | |
| "grad_norm": 0.4503066837787628, | |
| "learning_rate": 0.00030559464181712285, | |
| "loss": 3.1933, | |
| "step": 84350 | |
| }, | |
| { | |
| "epoch": 24.56357708430368, | |
| "grad_norm": 0.4374663233757019, | |
| "learning_rate": 0.0003054199184624345, | |
| "loss": 3.1984, | |
| "step": 84400 | |
| }, | |
| { | |
| "epoch": 24.578132277596648, | |
| "grad_norm": 0.42423900961875916, | |
| "learning_rate": 0.00030524519510774604, | |
| "loss": 3.1828, | |
| "step": 84450 | |
| }, | |
| { | |
| "epoch": 24.592687470889615, | |
| "grad_norm": 0.4310375452041626, | |
| "learning_rate": 0.00030507047175305763, | |
| "loss": 3.1914, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 24.60724266418258, | |
| "grad_norm": 0.3991932272911072, | |
| "learning_rate": 0.0003048957483983692, | |
| "loss": 3.1917, | |
| "step": 84550 | |
| }, | |
| { | |
| "epoch": 24.621797857475546, | |
| "grad_norm": 0.4246843457221985, | |
| "learning_rate": 0.00030472102504368076, | |
| "loss": 3.2037, | |
| "step": 84600 | |
| }, | |
| { | |
| "epoch": 24.636353050768513, | |
| "grad_norm": 0.4228966236114502, | |
| "learning_rate": 0.0003045463016889924, | |
| "loss": 3.1958, | |
| "step": 84650 | |
| }, | |
| { | |
| "epoch": 24.65090824406148, | |
| "grad_norm": 0.40659230947494507, | |
| "learning_rate": 0.000304371578334304, | |
| "loss": 3.2011, | |
| "step": 84700 | |
| }, | |
| { | |
| "epoch": 24.665463437354447, | |
| "grad_norm": 0.42217251658439636, | |
| "learning_rate": 0.0003041968549796156, | |
| "loss": 3.195, | |
| "step": 84750 | |
| }, | |
| { | |
| "epoch": 24.680018630647414, | |
| "grad_norm": 0.4322725832462311, | |
| "learning_rate": 0.00030402213162492714, | |
| "loss": 3.1958, | |
| "step": 84800 | |
| }, | |
| { | |
| "epoch": 24.69457382394038, | |
| "grad_norm": 0.43233662843704224, | |
| "learning_rate": 0.00030384740827023874, | |
| "loss": 3.1984, | |
| "step": 84850 | |
| }, | |
| { | |
| "epoch": 24.709129017233348, | |
| "grad_norm": 0.4197286069393158, | |
| "learning_rate": 0.00030367268491555033, | |
| "loss": 3.1842, | |
| "step": 84900 | |
| }, | |
| { | |
| "epoch": 24.723684210526315, | |
| "grad_norm": 0.42754897475242615, | |
| "learning_rate": 0.000303497961560862, | |
| "loss": 3.1976, | |
| "step": 84950 | |
| }, | |
| { | |
| "epoch": 24.738239403819282, | |
| "grad_norm": 0.41586819291114807, | |
| "learning_rate": 0.00030332323820617357, | |
| "loss": 3.2123, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.738239403819282, | |
| "eval_accuracy": 0.37482548664237536, | |
| "eval_loss": 3.5371224880218506, | |
| "eval_runtime": 82.8444, | |
| "eval_samples_per_second": 200.991, | |
| "eval_steps_per_second": 12.566, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.75279459711225, | |
| "grad_norm": 0.41994109749794006, | |
| "learning_rate": 0.0003031485148514851, | |
| "loss": 3.2076, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 24.767349790405216, | |
| "grad_norm": 0.44972968101501465, | |
| "learning_rate": 0.0003029737914967967, | |
| "loss": 3.2134, | |
| "step": 85100 | |
| }, | |
| { | |
| "epoch": 24.781904983698183, | |
| "grad_norm": 0.4217878580093384, | |
| "learning_rate": 0.0003027990681421083, | |
| "loss": 3.2031, | |
| "step": 85150 | |
| }, | |
| { | |
| "epoch": 24.79646017699115, | |
| "grad_norm": 0.45532307028770447, | |
| "learning_rate": 0.00030262434478741984, | |
| "loss": 3.2051, | |
| "step": 85200 | |
| }, | |
| { | |
| "epoch": 24.811015370284117, | |
| "grad_norm": 0.4029344916343689, | |
| "learning_rate": 0.0003024496214327315, | |
| "loss": 3.207, | |
| "step": 85250 | |
| }, | |
| { | |
| "epoch": 24.825570563577084, | |
| "grad_norm": 0.44731462001800537, | |
| "learning_rate": 0.0003022748980780431, | |
| "loss": 3.1836, | |
| "step": 85300 | |
| }, | |
| { | |
| "epoch": 24.84012575687005, | |
| "grad_norm": 0.418901652097702, | |
| "learning_rate": 0.0003021001747233547, | |
| "loss": 3.2191, | |
| "step": 85350 | |
| }, | |
| { | |
| "epoch": 24.85468095016302, | |
| "grad_norm": 0.45161646604537964, | |
| "learning_rate": 0.0003019254513686662, | |
| "loss": 3.2067, | |
| "step": 85400 | |
| }, | |
| { | |
| "epoch": 24.869236143455986, | |
| "grad_norm": 0.43810057640075684, | |
| "learning_rate": 0.0003017507280139778, | |
| "loss": 3.2176, | |
| "step": 85450 | |
| }, | |
| { | |
| "epoch": 24.883791336748953, | |
| "grad_norm": 0.4479285776615143, | |
| "learning_rate": 0.00030157600465928946, | |
| "loss": 3.2088, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 24.89834653004192, | |
| "grad_norm": 0.42635491490364075, | |
| "learning_rate": 0.00030140128130460106, | |
| "loss": 3.2174, | |
| "step": 85550 | |
| }, | |
| { | |
| "epoch": 24.912901723334887, | |
| "grad_norm": 0.4362599849700928, | |
| "learning_rate": 0.0003012265579499126, | |
| "loss": 3.2091, | |
| "step": 85600 | |
| }, | |
| { | |
| "epoch": 24.927456916627854, | |
| "grad_norm": 0.43327707052230835, | |
| "learning_rate": 0.0003010518345952242, | |
| "loss": 3.2146, | |
| "step": 85650 | |
| }, | |
| { | |
| "epoch": 24.94201210992082, | |
| "grad_norm": 0.42312800884246826, | |
| "learning_rate": 0.0003008771112405358, | |
| "loss": 3.2136, | |
| "step": 85700 | |
| }, | |
| { | |
| "epoch": 24.956567303213788, | |
| "grad_norm": 0.4401114583015442, | |
| "learning_rate": 0.0003007023878858473, | |
| "loss": 3.2153, | |
| "step": 85750 | |
| }, | |
| { | |
| "epoch": 24.971122496506755, | |
| "grad_norm": 0.4146105945110321, | |
| "learning_rate": 0.000300527664531159, | |
| "loss": 3.2052, | |
| "step": 85800 | |
| }, | |
| { | |
| "epoch": 24.985677689799722, | |
| "grad_norm": 0.4527598023414612, | |
| "learning_rate": 0.00030035294117647057, | |
| "loss": 3.2132, | |
| "step": 85850 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00030017821782178216, | |
| "loss": 3.221, | |
| "step": 85900 | |
| }, | |
| { | |
| "epoch": 25.014555193292967, | |
| "grad_norm": 0.4331722557544708, | |
| "learning_rate": 0.00030000349446709376, | |
| "loss": 3.1089, | |
| "step": 85950 | |
| }, | |
| { | |
| "epoch": 25.029110386585934, | |
| "grad_norm": 0.42661145329475403, | |
| "learning_rate": 0.00029982877111240535, | |
| "loss": 3.1188, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.029110386585934, | |
| "eval_accuracy": 0.37422821287626806, | |
| "eval_loss": 3.547804832458496, | |
| "eval_runtime": 82.8742, | |
| "eval_samples_per_second": 200.919, | |
| "eval_steps_per_second": 12.561, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.0436655798789, | |
| "grad_norm": 0.432282418012619, | |
| "learning_rate": 0.00029965404775771694, | |
| "loss": 3.1189, | |
| "step": 86050 | |
| }, | |
| { | |
| "epoch": 25.05822077317187, | |
| "grad_norm": 0.4282018542289734, | |
| "learning_rate": 0.0002994793244030285, | |
| "loss": 3.1266, | |
| "step": 86100 | |
| }, | |
| { | |
| "epoch": 25.072775966464835, | |
| "grad_norm": 0.4331301152706146, | |
| "learning_rate": 0.00029930460104834013, | |
| "loss": 3.1262, | |
| "step": 86150 | |
| }, | |
| { | |
| "epoch": 25.087331159757802, | |
| "grad_norm": 0.4194384515285492, | |
| "learning_rate": 0.0002991298776936517, | |
| "loss": 3.1292, | |
| "step": 86200 | |
| }, | |
| { | |
| "epoch": 25.10188635305077, | |
| "grad_norm": 0.45870241522789, | |
| "learning_rate": 0.00029895515433896327, | |
| "loss": 3.1404, | |
| "step": 86250 | |
| }, | |
| { | |
| "epoch": 25.116441546343736, | |
| "grad_norm": 0.41547247767448425, | |
| "learning_rate": 0.00029878043098427486, | |
| "loss": 3.134, | |
| "step": 86300 | |
| }, | |
| { | |
| "epoch": 25.130996739636704, | |
| "grad_norm": 0.4174516797065735, | |
| "learning_rate": 0.00029860570762958646, | |
| "loss": 3.1433, | |
| "step": 86350 | |
| }, | |
| { | |
| "epoch": 25.14555193292967, | |
| "grad_norm": 0.4430071711540222, | |
| "learning_rate": 0.00029843098427489805, | |
| "loss": 3.1416, | |
| "step": 86400 | |
| }, | |
| { | |
| "epoch": 25.160107126222638, | |
| "grad_norm": 0.40706923604011536, | |
| "learning_rate": 0.00029825626092020964, | |
| "loss": 3.1386, | |
| "step": 86450 | |
| }, | |
| { | |
| "epoch": 25.174662319515605, | |
| "grad_norm": 0.42693978548049927, | |
| "learning_rate": 0.00029808153756552124, | |
| "loss": 3.1383, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 25.189217512808572, | |
| "grad_norm": 0.4446597695350647, | |
| "learning_rate": 0.00029790681421083283, | |
| "loss": 3.1525, | |
| "step": 86550 | |
| }, | |
| { | |
| "epoch": 25.203772706101535, | |
| "grad_norm": 0.418632835149765, | |
| "learning_rate": 0.00029773209085614443, | |
| "loss": 3.1431, | |
| "step": 86600 | |
| }, | |
| { | |
| "epoch": 25.218327899394502, | |
| "grad_norm": 0.43115079402923584, | |
| "learning_rate": 0.00029755736750145597, | |
| "loss": 3.1449, | |
| "step": 86650 | |
| }, | |
| { | |
| "epoch": 25.23288309268747, | |
| "grad_norm": 0.44214192032814026, | |
| "learning_rate": 0.0002973826441467676, | |
| "loss": 3.1436, | |
| "step": 86700 | |
| }, | |
| { | |
| "epoch": 25.247438285980436, | |
| "grad_norm": 0.42377567291259766, | |
| "learning_rate": 0.00029720792079207916, | |
| "loss": 3.1557, | |
| "step": 86750 | |
| }, | |
| { | |
| "epoch": 25.261993479273404, | |
| "grad_norm": 0.4016624987125397, | |
| "learning_rate": 0.00029703319743739075, | |
| "loss": 3.1513, | |
| "step": 86800 | |
| }, | |
| { | |
| "epoch": 25.27654867256637, | |
| "grad_norm": 0.4337550103664398, | |
| "learning_rate": 0.00029685847408270234, | |
| "loss": 3.1636, | |
| "step": 86850 | |
| }, | |
| { | |
| "epoch": 25.291103865859338, | |
| "grad_norm": 0.4288932979106903, | |
| "learning_rate": 0.00029668375072801394, | |
| "loss": 3.1593, | |
| "step": 86900 | |
| }, | |
| { | |
| "epoch": 25.305659059152305, | |
| "grad_norm": 0.4708848297595978, | |
| "learning_rate": 0.00029650902737332553, | |
| "loss": 3.1557, | |
| "step": 86950 | |
| }, | |
| { | |
| "epoch": 25.320214252445272, | |
| "grad_norm": 0.41990965604782104, | |
| "learning_rate": 0.00029633430401863713, | |
| "loss": 3.1648, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.320214252445272, | |
| "eval_accuracy": 0.37435702280300037, | |
| "eval_loss": 3.545494794845581, | |
| "eval_runtime": 82.7922, | |
| "eval_samples_per_second": 201.118, | |
| "eval_steps_per_second": 12.574, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.33476944573824, | |
| "grad_norm": 0.45416632294654846, | |
| "learning_rate": 0.0002961595806639487, | |
| "loss": 3.1475, | |
| "step": 87050 | |
| }, | |
| { | |
| "epoch": 25.349324639031206, | |
| "grad_norm": 0.4437417685985565, | |
| "learning_rate": 0.0002959848573092603, | |
| "loss": 3.1689, | |
| "step": 87100 | |
| }, | |
| { | |
| "epoch": 25.363879832324173, | |
| "grad_norm": 0.42698612809181213, | |
| "learning_rate": 0.0002958101339545719, | |
| "loss": 3.1837, | |
| "step": 87150 | |
| }, | |
| { | |
| "epoch": 25.37843502561714, | |
| "grad_norm": 0.4282931685447693, | |
| "learning_rate": 0.0002956354105998835, | |
| "loss": 3.1686, | |
| "step": 87200 | |
| }, | |
| { | |
| "epoch": 25.392990218910107, | |
| "grad_norm": 0.42175963521003723, | |
| "learning_rate": 0.0002954606872451951, | |
| "loss": 3.1657, | |
| "step": 87250 | |
| }, | |
| { | |
| "epoch": 25.407545412203074, | |
| "grad_norm": 0.4243347644805908, | |
| "learning_rate": 0.0002952859638905067, | |
| "loss": 3.1643, | |
| "step": 87300 | |
| }, | |
| { | |
| "epoch": 25.42210060549604, | |
| "grad_norm": 0.4498177170753479, | |
| "learning_rate": 0.00029511124053581823, | |
| "loss": 3.17, | |
| "step": 87350 | |
| }, | |
| { | |
| "epoch": 25.43665579878901, | |
| "grad_norm": 0.4397185444831848, | |
| "learning_rate": 0.0002949365171811299, | |
| "loss": 3.1711, | |
| "step": 87400 | |
| }, | |
| { | |
| "epoch": 25.451210992081975, | |
| "grad_norm": 0.4264979064464569, | |
| "learning_rate": 0.0002947617938264414, | |
| "loss": 3.1657, | |
| "step": 87450 | |
| }, | |
| { | |
| "epoch": 25.465766185374942, | |
| "grad_norm": 0.4243890345096588, | |
| "learning_rate": 0.000294587070471753, | |
| "loss": 3.1713, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 25.48032137866791, | |
| "grad_norm": 0.4476009011268616, | |
| "learning_rate": 0.0002944123471170646, | |
| "loss": 3.1851, | |
| "step": 87550 | |
| }, | |
| { | |
| "epoch": 25.494876571960877, | |
| "grad_norm": 0.4547466039657593, | |
| "learning_rate": 0.0002942376237623762, | |
| "loss": 3.1975, | |
| "step": 87600 | |
| }, | |
| { | |
| "epoch": 25.509431765253844, | |
| "grad_norm": 0.41021642088890076, | |
| "learning_rate": 0.0002940629004076878, | |
| "loss": 3.1798, | |
| "step": 87650 | |
| }, | |
| { | |
| "epoch": 25.52398695854681, | |
| "grad_norm": 0.47481679916381836, | |
| "learning_rate": 0.0002938881770529994, | |
| "loss": 3.1791, | |
| "step": 87700 | |
| }, | |
| { | |
| "epoch": 25.538542151839778, | |
| "grad_norm": 0.4135197103023529, | |
| "learning_rate": 0.000293713453698311, | |
| "loss": 3.1843, | |
| "step": 87750 | |
| }, | |
| { | |
| "epoch": 25.553097345132745, | |
| "grad_norm": 0.4334968328475952, | |
| "learning_rate": 0.0002935387303436226, | |
| "loss": 3.1762, | |
| "step": 87800 | |
| }, | |
| { | |
| "epoch": 25.567652538425712, | |
| "grad_norm": 0.4079611897468567, | |
| "learning_rate": 0.0002933640069889342, | |
| "loss": 3.1833, | |
| "step": 87850 | |
| }, | |
| { | |
| "epoch": 25.58220773171868, | |
| "grad_norm": 0.44633978605270386, | |
| "learning_rate": 0.0002931892836342457, | |
| "loss": 3.1871, | |
| "step": 87900 | |
| }, | |
| { | |
| "epoch": 25.596762925011646, | |
| "grad_norm": 0.4224945604801178, | |
| "learning_rate": 0.00029301456027955736, | |
| "loss": 3.1849, | |
| "step": 87950 | |
| }, | |
| { | |
| "epoch": 25.61131811830461, | |
| "grad_norm": 0.431976318359375, | |
| "learning_rate": 0.0002928398369248689, | |
| "loss": 3.1804, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.61131811830461, | |
| "eval_accuracy": 0.3748529880318419, | |
| "eval_loss": 3.5384273529052734, | |
| "eval_runtime": 82.8599, | |
| "eval_samples_per_second": 200.954, | |
| "eval_steps_per_second": 12.563, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.625873311597577, | |
| "grad_norm": 0.4322817325592041, | |
| "learning_rate": 0.0002926651135701805, | |
| "loss": 3.182, | |
| "step": 88050 | |
| }, | |
| { | |
| "epoch": 25.640428504890544, | |
| "grad_norm": 0.4342901408672333, | |
| "learning_rate": 0.00029249039021549215, | |
| "loss": 3.1869, | |
| "step": 88100 | |
| }, | |
| { | |
| "epoch": 25.65498369818351, | |
| "grad_norm": 0.4451529085636139, | |
| "learning_rate": 0.0002923156668608037, | |
| "loss": 3.1896, | |
| "step": 88150 | |
| }, | |
| { | |
| "epoch": 25.669538891476478, | |
| "grad_norm": 0.4365444779396057, | |
| "learning_rate": 0.0002921409435061153, | |
| "loss": 3.185, | |
| "step": 88200 | |
| }, | |
| { | |
| "epoch": 25.684094084769445, | |
| "grad_norm": 0.4128621816635132, | |
| "learning_rate": 0.0002919662201514269, | |
| "loss": 3.1835, | |
| "step": 88250 | |
| }, | |
| { | |
| "epoch": 25.698649278062412, | |
| "grad_norm": 0.4501856565475464, | |
| "learning_rate": 0.00029179149679673847, | |
| "loss": 3.1821, | |
| "step": 88300 | |
| }, | |
| { | |
| "epoch": 25.71320447135538, | |
| "grad_norm": 0.4202530086040497, | |
| "learning_rate": 0.00029161677344205007, | |
| "loss": 3.188, | |
| "step": 88350 | |
| }, | |
| { | |
| "epoch": 25.727759664648346, | |
| "grad_norm": 0.4299803078174591, | |
| "learning_rate": 0.00029144205008736166, | |
| "loss": 3.1986, | |
| "step": 88400 | |
| }, | |
| { | |
| "epoch": 25.742314857941313, | |
| "grad_norm": 0.4290253221988678, | |
| "learning_rate": 0.00029126732673267325, | |
| "loss": 3.1996, | |
| "step": 88450 | |
| }, | |
| { | |
| "epoch": 25.75687005123428, | |
| "grad_norm": 0.42774584889411926, | |
| "learning_rate": 0.00029109260337798485, | |
| "loss": 3.1947, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 25.771425244527247, | |
| "grad_norm": 0.4753718078136444, | |
| "learning_rate": 0.00029091788002329644, | |
| "loss": 3.1883, | |
| "step": 88550 | |
| }, | |
| { | |
| "epoch": 25.785980437820214, | |
| "grad_norm": 0.4146498441696167, | |
| "learning_rate": 0.000290743156668608, | |
| "loss": 3.1919, | |
| "step": 88600 | |
| }, | |
| { | |
| "epoch": 25.80053563111318, | |
| "grad_norm": 0.455722838640213, | |
| "learning_rate": 0.00029056843331391963, | |
| "loss": 3.2051, | |
| "step": 88650 | |
| }, | |
| { | |
| "epoch": 25.81509082440615, | |
| "grad_norm": 0.4132815897464752, | |
| "learning_rate": 0.00029039370995923117, | |
| "loss": 3.1925, | |
| "step": 88700 | |
| }, | |
| { | |
| "epoch": 25.829646017699115, | |
| "grad_norm": 0.41768813133239746, | |
| "learning_rate": 0.00029021898660454277, | |
| "loss": 3.2009, | |
| "step": 88750 | |
| }, | |
| { | |
| "epoch": 25.844201210992082, | |
| "grad_norm": 0.43633952736854553, | |
| "learning_rate": 0.00029004426324985436, | |
| "loss": 3.1981, | |
| "step": 88800 | |
| }, | |
| { | |
| "epoch": 25.85875640428505, | |
| "grad_norm": 0.443166047334671, | |
| "learning_rate": 0.00028986953989516595, | |
| "loss": 3.1971, | |
| "step": 88850 | |
| }, | |
| { | |
| "epoch": 25.873311597578017, | |
| "grad_norm": 0.413812518119812, | |
| "learning_rate": 0.00028969481654047755, | |
| "loss": 3.2019, | |
| "step": 88900 | |
| }, | |
| { | |
| "epoch": 25.887866790870984, | |
| "grad_norm": 0.41359344124794006, | |
| "learning_rate": 0.00028952009318578914, | |
| "loss": 3.2079, | |
| "step": 88950 | |
| }, | |
| { | |
| "epoch": 25.90242198416395, | |
| "grad_norm": 0.43147844076156616, | |
| "learning_rate": 0.00028934536983110074, | |
| "loss": 3.2063, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.90242198416395, | |
| "eval_accuracy": 0.37529336284522324, | |
| "eval_loss": 3.531651735305786, | |
| "eval_runtime": 82.8248, | |
| "eval_samples_per_second": 201.039, | |
| "eval_steps_per_second": 12.569, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.916977177456918, | |
| "grad_norm": 0.4183621406555176, | |
| "learning_rate": 0.00028917064647641233, | |
| "loss": 3.2147, | |
| "step": 89050 | |
| }, | |
| { | |
| "epoch": 25.931532370749885, | |
| "grad_norm": 0.4477531909942627, | |
| "learning_rate": 0.0002889959231217239, | |
| "loss": 3.2097, | |
| "step": 89100 | |
| }, | |
| { | |
| "epoch": 25.946087564042852, | |
| "grad_norm": 0.4175955355167389, | |
| "learning_rate": 0.0002888211997670355, | |
| "loss": 3.192, | |
| "step": 89150 | |
| }, | |
| { | |
| "epoch": 25.96064275733582, | |
| "grad_norm": 0.4285762906074524, | |
| "learning_rate": 0.0002886464764123471, | |
| "loss": 3.2052, | |
| "step": 89200 | |
| }, | |
| { | |
| "epoch": 25.975197950628786, | |
| "grad_norm": 0.4780668318271637, | |
| "learning_rate": 0.0002884717530576587, | |
| "loss": 3.2023, | |
| "step": 89250 | |
| }, | |
| { | |
| "epoch": 25.989753143921753, | |
| "grad_norm": 0.45964622497558594, | |
| "learning_rate": 0.00028829702970297025, | |
| "loss": 3.1928, | |
| "step": 89300 | |
| }, | |
| { | |
| "epoch": 26.00407545412203, | |
| "grad_norm": 0.4359065890312195, | |
| "learning_rate": 0.0002881223063482819, | |
| "loss": 3.164, | |
| "step": 89350 | |
| }, | |
| { | |
| "epoch": 26.018630647414998, | |
| "grad_norm": 0.4142531752586365, | |
| "learning_rate": 0.00028794758299359344, | |
| "loss": 3.1033, | |
| "step": 89400 | |
| }, | |
| { | |
| "epoch": 26.033185840707965, | |
| "grad_norm": 0.4193040132522583, | |
| "learning_rate": 0.00028777285963890503, | |
| "loss": 3.1077, | |
| "step": 89450 | |
| }, | |
| { | |
| "epoch": 26.047741034000932, | |
| "grad_norm": 0.4525713324546814, | |
| "learning_rate": 0.0002875981362842166, | |
| "loss": 3.1237, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 26.0622962272939, | |
| "grad_norm": 0.42168205976486206, | |
| "learning_rate": 0.0002874234129295282, | |
| "loss": 3.121, | |
| "step": 89550 | |
| }, | |
| { | |
| "epoch": 26.076851420586866, | |
| "grad_norm": 0.4232228994369507, | |
| "learning_rate": 0.0002872486895748398, | |
| "loss": 3.1299, | |
| "step": 89600 | |
| }, | |
| { | |
| "epoch": 26.091406613879833, | |
| "grad_norm": 0.4403054714202881, | |
| "learning_rate": 0.0002870739662201514, | |
| "loss": 3.1298, | |
| "step": 89650 | |
| }, | |
| { | |
| "epoch": 26.1059618071728, | |
| "grad_norm": 0.41803574562072754, | |
| "learning_rate": 0.000286899242865463, | |
| "loss": 3.1163, | |
| "step": 89700 | |
| }, | |
| { | |
| "epoch": 26.120517000465767, | |
| "grad_norm": 0.4274555742740631, | |
| "learning_rate": 0.0002867245195107746, | |
| "loss": 3.1212, | |
| "step": 89750 | |
| }, | |
| { | |
| "epoch": 26.135072193758734, | |
| "grad_norm": 0.46480271220207214, | |
| "learning_rate": 0.0002865497961560862, | |
| "loss": 3.1263, | |
| "step": 89800 | |
| }, | |
| { | |
| "epoch": 26.1496273870517, | |
| "grad_norm": 0.4393923580646515, | |
| "learning_rate": 0.00028637507280139773, | |
| "loss": 3.1338, | |
| "step": 89850 | |
| }, | |
| { | |
| "epoch": 26.16418258034467, | |
| "grad_norm": 0.42707085609436035, | |
| "learning_rate": 0.0002862003494467094, | |
| "loss": 3.1334, | |
| "step": 89900 | |
| }, | |
| { | |
| "epoch": 26.178737773637636, | |
| "grad_norm": 0.4436829388141632, | |
| "learning_rate": 0.0002860256260920209, | |
| "loss": 3.1396, | |
| "step": 89950 | |
| }, | |
| { | |
| "epoch": 26.1932929669306, | |
| "grad_norm": 0.4210873544216156, | |
| "learning_rate": 0.0002858509027373325, | |
| "loss": 3.1342, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.1932929669306, | |
| "eval_accuracy": 0.3744360011522377, | |
| "eval_loss": 3.5479485988616943, | |
| "eval_runtime": 82.825, | |
| "eval_samples_per_second": 201.038, | |
| "eval_steps_per_second": 12.569, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.207848160223566, | |
| "grad_norm": 0.449847012758255, | |
| "learning_rate": 0.0002856761793826441, | |
| "loss": 3.1379, | |
| "step": 90050 | |
| }, | |
| { | |
| "epoch": 26.222403353516533, | |
| "grad_norm": 0.4338505268096924, | |
| "learning_rate": 0.0002855014560279557, | |
| "loss": 3.1329, | |
| "step": 90100 | |
| }, | |
| { | |
| "epoch": 26.2369585468095, | |
| "grad_norm": 0.41904863715171814, | |
| "learning_rate": 0.0002853267326732673, | |
| "loss": 3.1416, | |
| "step": 90150 | |
| }, | |
| { | |
| "epoch": 26.251513740102467, | |
| "grad_norm": 0.4501389265060425, | |
| "learning_rate": 0.0002851520093185789, | |
| "loss": 3.1402, | |
| "step": 90200 | |
| }, | |
| { | |
| "epoch": 26.266068933395434, | |
| "grad_norm": 0.4450111985206604, | |
| "learning_rate": 0.0002849772859638905, | |
| "loss": 3.153, | |
| "step": 90250 | |
| }, | |
| { | |
| "epoch": 26.2806241266884, | |
| "grad_norm": 0.42784583568573, | |
| "learning_rate": 0.0002848025626092021, | |
| "loss": 3.1366, | |
| "step": 90300 | |
| }, | |
| { | |
| "epoch": 26.29517931998137, | |
| "grad_norm": 0.4506232738494873, | |
| "learning_rate": 0.0002846278392545137, | |
| "loss": 3.1519, | |
| "step": 90350 | |
| }, | |
| { | |
| "epoch": 26.309734513274336, | |
| "grad_norm": 0.4308539032936096, | |
| "learning_rate": 0.00028445311589982527, | |
| "loss": 3.1349, | |
| "step": 90400 | |
| }, | |
| { | |
| "epoch": 26.324289706567303, | |
| "grad_norm": 0.47364309430122375, | |
| "learning_rate": 0.00028427839254513686, | |
| "loss": 3.1517, | |
| "step": 90450 | |
| }, | |
| { | |
| "epoch": 26.33884489986027, | |
| "grad_norm": 0.4357488453388214, | |
| "learning_rate": 0.00028410366919044846, | |
| "loss": 3.1588, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 26.353400093153237, | |
| "grad_norm": 0.43206942081451416, | |
| "learning_rate": 0.00028392894583576, | |
| "loss": 3.1532, | |
| "step": 90550 | |
| }, | |
| { | |
| "epoch": 26.367955286446204, | |
| "grad_norm": 0.4371725916862488, | |
| "learning_rate": 0.00028375422248107165, | |
| "loss": 3.1484, | |
| "step": 90600 | |
| }, | |
| { | |
| "epoch": 26.38251047973917, | |
| "grad_norm": 0.4228227138519287, | |
| "learning_rate": 0.0002835794991263832, | |
| "loss": 3.1701, | |
| "step": 90650 | |
| }, | |
| { | |
| "epoch": 26.397065673032138, | |
| "grad_norm": 0.4390973746776581, | |
| "learning_rate": 0.0002834047757716948, | |
| "loss": 3.1763, | |
| "step": 90700 | |
| }, | |
| { | |
| "epoch": 26.411620866325105, | |
| "grad_norm": 0.4437226355075836, | |
| "learning_rate": 0.0002832300524170064, | |
| "loss": 3.1639, | |
| "step": 90750 | |
| }, | |
| { | |
| "epoch": 26.426176059618072, | |
| "grad_norm": 0.42696788907051086, | |
| "learning_rate": 0.00028305532906231797, | |
| "loss": 3.172, | |
| "step": 90800 | |
| }, | |
| { | |
| "epoch": 26.44073125291104, | |
| "grad_norm": 0.4321151077747345, | |
| "learning_rate": 0.00028288060570762956, | |
| "loss": 3.1663, | |
| "step": 90850 | |
| }, | |
| { | |
| "epoch": 26.455286446204006, | |
| "grad_norm": 0.44679439067840576, | |
| "learning_rate": 0.00028270588235294116, | |
| "loss": 3.1665, | |
| "step": 90900 | |
| }, | |
| { | |
| "epoch": 26.469841639496973, | |
| "grad_norm": 0.4245719909667969, | |
| "learning_rate": 0.00028253115899825275, | |
| "loss": 3.1726, | |
| "step": 90950 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "grad_norm": 0.44278499484062195, | |
| "learning_rate": 0.0002823564356435643, | |
| "loss": 3.1588, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "eval_accuracy": 0.3749048175735289, | |
| "eval_loss": 3.5422563552856445, | |
| "eval_runtime": 82.8325, | |
| "eval_samples_per_second": 201.02, | |
| "eval_steps_per_second": 12.568, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.498952026082907, | |
| "grad_norm": 0.4553212821483612, | |
| "learning_rate": 0.00028218171228887594, | |
| "loss": 3.1707, | |
| "step": 91050 | |
| }, | |
| { | |
| "epoch": 26.513507219375875, | |
| "grad_norm": 0.450360506772995, | |
| "learning_rate": 0.0002820069889341875, | |
| "loss": 3.1581, | |
| "step": 91100 | |
| }, | |
| { | |
| "epoch": 26.52806241266884, | |
| "grad_norm": 0.4682866930961609, | |
| "learning_rate": 0.00028183226557949913, | |
| "loss": 3.176, | |
| "step": 91150 | |
| }, | |
| { | |
| "epoch": 26.54261760596181, | |
| "grad_norm": 0.44312798976898193, | |
| "learning_rate": 0.00028165754222481067, | |
| "loss": 3.1787, | |
| "step": 91200 | |
| }, | |
| { | |
| "epoch": 26.557172799254776, | |
| "grad_norm": 0.4540131390094757, | |
| "learning_rate": 0.00028148281887012226, | |
| "loss": 3.1769, | |
| "step": 91250 | |
| }, | |
| { | |
| "epoch": 26.571727992547743, | |
| "grad_norm": 0.4430559277534485, | |
| "learning_rate": 0.0002813080955154339, | |
| "loss": 3.1699, | |
| "step": 91300 | |
| }, | |
| { | |
| "epoch": 26.586283185840706, | |
| "grad_norm": 0.4341665804386139, | |
| "learning_rate": 0.00028113337216074545, | |
| "loss": 3.1812, | |
| "step": 91350 | |
| }, | |
| { | |
| "epoch": 26.600838379133673, | |
| "grad_norm": 0.4232028126716614, | |
| "learning_rate": 0.00028095864880605705, | |
| "loss": 3.1835, | |
| "step": 91400 | |
| }, | |
| { | |
| "epoch": 26.61539357242664, | |
| "grad_norm": 0.42725902795791626, | |
| "learning_rate": 0.00028078392545136864, | |
| "loss": 3.1702, | |
| "step": 91450 | |
| }, | |
| { | |
| "epoch": 26.629948765719607, | |
| "grad_norm": 0.4316592812538147, | |
| "learning_rate": 0.00028060920209668023, | |
| "loss": 3.1758, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 26.644503959012575, | |
| "grad_norm": 0.4443933367729187, | |
| "learning_rate": 0.00028043447874199183, | |
| "loss": 3.1649, | |
| "step": 91550 | |
| }, | |
| { | |
| "epoch": 26.65905915230554, | |
| "grad_norm": 0.4466203451156616, | |
| "learning_rate": 0.0002802597553873034, | |
| "loss": 3.1726, | |
| "step": 91600 | |
| }, | |
| { | |
| "epoch": 26.67361434559851, | |
| "grad_norm": 0.45013076066970825, | |
| "learning_rate": 0.000280085032032615, | |
| "loss": 3.177, | |
| "step": 91650 | |
| }, | |
| { | |
| "epoch": 26.688169538891476, | |
| "grad_norm": 0.4297407567501068, | |
| "learning_rate": 0.00027991030867792656, | |
| "loss": 3.18, | |
| "step": 91700 | |
| }, | |
| { | |
| "epoch": 26.702724732184443, | |
| "grad_norm": 0.44508230686187744, | |
| "learning_rate": 0.0002797355853232382, | |
| "loss": 3.177, | |
| "step": 91750 | |
| }, | |
| { | |
| "epoch": 26.71727992547741, | |
| "grad_norm": 0.48300066590309143, | |
| "learning_rate": 0.00027956086196854975, | |
| "loss": 3.1859, | |
| "step": 91800 | |
| }, | |
| { | |
| "epoch": 26.731835118770377, | |
| "grad_norm": 0.4367406666278839, | |
| "learning_rate": 0.0002793861386138614, | |
| "loss": 3.2024, | |
| "step": 91850 | |
| }, | |
| { | |
| "epoch": 26.746390312063344, | |
| "grad_norm": 0.46090418100357056, | |
| "learning_rate": 0.00027921141525917293, | |
| "loss": 3.1909, | |
| "step": 91900 | |
| }, | |
| { | |
| "epoch": 26.76094550535631, | |
| "grad_norm": 0.45331886410713196, | |
| "learning_rate": 0.00027903669190448453, | |
| "loss": 3.1827, | |
| "step": 91950 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "grad_norm": 0.4293125569820404, | |
| "learning_rate": 0.0002788619685497961, | |
| "loss": 3.1962, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "eval_accuracy": 0.3752381250116793, | |
| "eval_loss": 3.5354275703430176, | |
| "eval_runtime": 82.7687, | |
| "eval_samples_per_second": 201.175, | |
| "eval_steps_per_second": 12.577, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.790055891942245, | |
| "grad_norm": 0.4252769351005554, | |
| "learning_rate": 0.0002786872451951077, | |
| "loss": 3.1873, | |
| "step": 92050 | |
| }, | |
| { | |
| "epoch": 26.804611085235212, | |
| "grad_norm": 0.4190332889556885, | |
| "learning_rate": 0.0002785125218404193, | |
| "loss": 3.1885, | |
| "step": 92100 | |
| }, | |
| { | |
| "epoch": 26.81916627852818, | |
| "grad_norm": 0.4223249852657318, | |
| "learning_rate": 0.0002783377984857309, | |
| "loss": 3.1906, | |
| "step": 92150 | |
| }, | |
| { | |
| "epoch": 26.833721471821146, | |
| "grad_norm": 0.44625985622406006, | |
| "learning_rate": 0.0002781630751310425, | |
| "loss": 3.1887, | |
| "step": 92200 | |
| }, | |
| { | |
| "epoch": 26.848276665114113, | |
| "grad_norm": 0.4521753191947937, | |
| "learning_rate": 0.0002779883517763541, | |
| "loss": 3.187, | |
| "step": 92250 | |
| }, | |
| { | |
| "epoch": 26.86283185840708, | |
| "grad_norm": 0.44966092705726624, | |
| "learning_rate": 0.0002778136284216657, | |
| "loss": 3.1927, | |
| "step": 92300 | |
| }, | |
| { | |
| "epoch": 26.877387051700047, | |
| "grad_norm": 0.4730190634727478, | |
| "learning_rate": 0.0002776389050669773, | |
| "loss": 3.1898, | |
| "step": 92350 | |
| }, | |
| { | |
| "epoch": 26.891942244993015, | |
| "grad_norm": 0.41821205615997314, | |
| "learning_rate": 0.0002774641817122888, | |
| "loss": 3.1841, | |
| "step": 92400 | |
| }, | |
| { | |
| "epoch": 26.90649743828598, | |
| "grad_norm": 0.441342294216156, | |
| "learning_rate": 0.00027728945835760047, | |
| "loss": 3.1934, | |
| "step": 92450 | |
| }, | |
| { | |
| "epoch": 26.92105263157895, | |
| "grad_norm": 0.44856083393096924, | |
| "learning_rate": 0.000277114735002912, | |
| "loss": 3.1962, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 26.935607824871916, | |
| "grad_norm": 0.43550050258636475, | |
| "learning_rate": 0.00027694001164822366, | |
| "loss": 3.1873, | |
| "step": 92550 | |
| }, | |
| { | |
| "epoch": 26.950163018164883, | |
| "grad_norm": 0.46068528294563293, | |
| "learning_rate": 0.0002767652882935352, | |
| "loss": 3.1932, | |
| "step": 92600 | |
| }, | |
| { | |
| "epoch": 26.96471821145785, | |
| "grad_norm": 0.4151133894920349, | |
| "learning_rate": 0.0002765905649388468, | |
| "loss": 3.187, | |
| "step": 92650 | |
| }, | |
| { | |
| "epoch": 26.979273404750813, | |
| "grad_norm": 0.43730881810188293, | |
| "learning_rate": 0.0002764158415841584, | |
| "loss": 3.2008, | |
| "step": 92700 | |
| }, | |
| { | |
| "epoch": 26.99382859804378, | |
| "grad_norm": 0.42695116996765137, | |
| "learning_rate": 0.00027624111822947, | |
| "loss": 3.1931, | |
| "step": 92750 | |
| }, | |
| { | |
| "epoch": 27.008150908244062, | |
| "grad_norm": 0.4479392468929291, | |
| "learning_rate": 0.0002760663948747816, | |
| "loss": 3.1385, | |
| "step": 92800 | |
| }, | |
| { | |
| "epoch": 27.02270610153703, | |
| "grad_norm": 0.4426802396774292, | |
| "learning_rate": 0.00027589167152009317, | |
| "loss": 3.1013, | |
| "step": 92850 | |
| }, | |
| { | |
| "epoch": 27.037261294829996, | |
| "grad_norm": 0.44218266010284424, | |
| "learning_rate": 0.00027571694816540477, | |
| "loss": 3.1045, | |
| "step": 92900 | |
| }, | |
| { | |
| "epoch": 27.051816488122963, | |
| "grad_norm": 0.472148597240448, | |
| "learning_rate": 0.0002755422248107163, | |
| "loss": 3.1093, | |
| "step": 92950 | |
| }, | |
| { | |
| "epoch": 27.06637168141593, | |
| "grad_norm": 0.4388134181499481, | |
| "learning_rate": 0.00027536750145602795, | |
| "loss": 3.1086, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.06637168141593, | |
| "eval_accuracy": 0.37412161561025875, | |
| "eval_loss": 3.5494494438171387, | |
| "eval_runtime": 82.8053, | |
| "eval_samples_per_second": 201.086, | |
| "eval_steps_per_second": 12.572, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.080926874708897, | |
| "grad_norm": 0.451478511095047, | |
| "learning_rate": 0.0002751927781013395, | |
| "loss": 3.1107, | |
| "step": 93050 | |
| }, | |
| { | |
| "epoch": 27.095482068001864, | |
| "grad_norm": 0.4257965385913849, | |
| "learning_rate": 0.0002750180547466511, | |
| "loss": 3.116, | |
| "step": 93100 | |
| }, | |
| { | |
| "epoch": 27.11003726129483, | |
| "grad_norm": 0.4503708481788635, | |
| "learning_rate": 0.0002748433313919627, | |
| "loss": 3.1166, | |
| "step": 93150 | |
| }, | |
| { | |
| "epoch": 27.1245924545878, | |
| "grad_norm": 0.4574490785598755, | |
| "learning_rate": 0.0002746686080372743, | |
| "loss": 3.1221, | |
| "step": 93200 | |
| }, | |
| { | |
| "epoch": 27.139147647880765, | |
| "grad_norm": 0.4546046555042267, | |
| "learning_rate": 0.00027449388468258587, | |
| "loss": 3.1323, | |
| "step": 93250 | |
| }, | |
| { | |
| "epoch": 27.153702841173732, | |
| "grad_norm": 0.44228723645210266, | |
| "learning_rate": 0.00027431916132789747, | |
| "loss": 3.1204, | |
| "step": 93300 | |
| }, | |
| { | |
| "epoch": 27.168258034466696, | |
| "grad_norm": 0.4699128270149231, | |
| "learning_rate": 0.00027414443797320906, | |
| "loss": 3.1288, | |
| "step": 93350 | |
| }, | |
| { | |
| "epoch": 27.182813227759663, | |
| "grad_norm": 0.44085732102394104, | |
| "learning_rate": 0.00027396971461852065, | |
| "loss": 3.1264, | |
| "step": 93400 | |
| }, | |
| { | |
| "epoch": 27.19736842105263, | |
| "grad_norm": 0.47711458802223206, | |
| "learning_rate": 0.00027379499126383225, | |
| "loss": 3.1349, | |
| "step": 93450 | |
| }, | |
| { | |
| "epoch": 27.211923614345597, | |
| "grad_norm": 0.46176978945732117, | |
| "learning_rate": 0.00027362026790914384, | |
| "loss": 3.1255, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 27.226478807638564, | |
| "grad_norm": 0.41149282455444336, | |
| "learning_rate": 0.00027344554455445544, | |
| "loss": 3.1347, | |
| "step": 93550 | |
| }, | |
| { | |
| "epoch": 27.24103400093153, | |
| "grad_norm": 0.4363878667354584, | |
| "learning_rate": 0.00027327082119976703, | |
| "loss": 3.1344, | |
| "step": 93600 | |
| }, | |
| { | |
| "epoch": 27.2555891942245, | |
| "grad_norm": 0.47142499685287476, | |
| "learning_rate": 0.00027309609784507857, | |
| "loss": 3.1449, | |
| "step": 93650 | |
| }, | |
| { | |
| "epoch": 27.270144387517465, | |
| "grad_norm": 0.43943706154823303, | |
| "learning_rate": 0.0002729213744903902, | |
| "loss": 3.1378, | |
| "step": 93700 | |
| }, | |
| { | |
| "epoch": 27.284699580810432, | |
| "grad_norm": 0.44916486740112305, | |
| "learning_rate": 0.00027274665113570176, | |
| "loss": 3.1454, | |
| "step": 93750 | |
| }, | |
| { | |
| "epoch": 27.2992547741034, | |
| "grad_norm": 0.43345949053764343, | |
| "learning_rate": 0.00027257192778101335, | |
| "loss": 3.1536, | |
| "step": 93800 | |
| }, | |
| { | |
| "epoch": 27.313809967396367, | |
| "grad_norm": 0.45956552028656006, | |
| "learning_rate": 0.00027239720442632495, | |
| "loss": 3.1318, | |
| "step": 93850 | |
| }, | |
| { | |
| "epoch": 27.328365160689334, | |
| "grad_norm": 0.44862690567970276, | |
| "learning_rate": 0.00027222248107163654, | |
| "loss": 3.1444, | |
| "step": 93900 | |
| }, | |
| { | |
| "epoch": 27.3429203539823, | |
| "grad_norm": 0.4102250933647156, | |
| "learning_rate": 0.00027204775771694814, | |
| "loss": 3.1343, | |
| "step": 93950 | |
| }, | |
| { | |
| "epoch": 27.357475547275268, | |
| "grad_norm": 0.42540040612220764, | |
| "learning_rate": 0.00027187303436225973, | |
| "loss": 3.1521, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.357475547275268, | |
| "eval_accuracy": 0.37496910500958963, | |
| "eval_loss": 3.544682741165161, | |
| "eval_runtime": 82.8208, | |
| "eval_samples_per_second": 201.049, | |
| "eval_steps_per_second": 12.569, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.372030740568235, | |
| "grad_norm": 0.43942588567733765, | |
| "learning_rate": 0.0002716983110075713, | |
| "loss": 3.1426, | |
| "step": 94050 | |
| }, | |
| { | |
| "epoch": 27.386585933861202, | |
| "grad_norm": 0.43729597330093384, | |
| "learning_rate": 0.0002715235876528829, | |
| "loss": 3.1525, | |
| "step": 94100 | |
| }, | |
| { | |
| "epoch": 27.40114112715417, | |
| "grad_norm": 0.4404613971710205, | |
| "learning_rate": 0.0002713488642981945, | |
| "loss": 3.1508, | |
| "step": 94150 | |
| }, | |
| { | |
| "epoch": 27.415696320447136, | |
| "grad_norm": 0.43753501772880554, | |
| "learning_rate": 0.00027117414094350606, | |
| "loss": 3.1588, | |
| "step": 94200 | |
| }, | |
| { | |
| "epoch": 27.430251513740103, | |
| "grad_norm": 0.44562819600105286, | |
| "learning_rate": 0.0002709994175888177, | |
| "loss": 3.1611, | |
| "step": 94250 | |
| }, | |
| { | |
| "epoch": 27.44480670703307, | |
| "grad_norm": 0.45446890592575073, | |
| "learning_rate": 0.00027082469423412924, | |
| "loss": 3.1559, | |
| "step": 94300 | |
| }, | |
| { | |
| "epoch": 27.459361900326037, | |
| "grad_norm": 0.45057451725006104, | |
| "learning_rate": 0.00027064997087944084, | |
| "loss": 3.1702, | |
| "step": 94350 | |
| }, | |
| { | |
| "epoch": 27.473917093619004, | |
| "grad_norm": 0.4548908770084381, | |
| "learning_rate": 0.00027047524752475243, | |
| "loss": 3.1553, | |
| "step": 94400 | |
| }, | |
| { | |
| "epoch": 27.48847228691197, | |
| "grad_norm": 0.4307977259159088, | |
| "learning_rate": 0.000270300524170064, | |
| "loss": 3.1659, | |
| "step": 94450 | |
| }, | |
| { | |
| "epoch": 27.50302748020494, | |
| "grad_norm": 0.46379518508911133, | |
| "learning_rate": 0.0002701258008153757, | |
| "loss": 3.1641, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 27.517582673497905, | |
| "grad_norm": 0.4659824073314667, | |
| "learning_rate": 0.0002699510774606872, | |
| "loss": 3.1566, | |
| "step": 94550 | |
| }, | |
| { | |
| "epoch": 27.532137866790872, | |
| "grad_norm": 0.4417530298233032, | |
| "learning_rate": 0.0002697763541059988, | |
| "loss": 3.173, | |
| "step": 94600 | |
| }, | |
| { | |
| "epoch": 27.54669306008384, | |
| "grad_norm": 0.4490537643432617, | |
| "learning_rate": 0.0002696016307513104, | |
| "loss": 3.1724, | |
| "step": 94650 | |
| }, | |
| { | |
| "epoch": 27.561248253376803, | |
| "grad_norm": 0.47829723358154297, | |
| "learning_rate": 0.000269426907396622, | |
| "loss": 3.162, | |
| "step": 94700 | |
| }, | |
| { | |
| "epoch": 27.57580344666977, | |
| "grad_norm": 0.47299665212631226, | |
| "learning_rate": 0.0002692521840419336, | |
| "loss": 3.1676, | |
| "step": 94750 | |
| }, | |
| { | |
| "epoch": 27.590358639962737, | |
| "grad_norm": 0.4622032046318054, | |
| "learning_rate": 0.0002690774606872452, | |
| "loss": 3.1602, | |
| "step": 94800 | |
| }, | |
| { | |
| "epoch": 27.604913833255704, | |
| "grad_norm": 0.4454992413520813, | |
| "learning_rate": 0.0002689027373325568, | |
| "loss": 3.1658, | |
| "step": 94850 | |
| }, | |
| { | |
| "epoch": 27.61946902654867, | |
| "grad_norm": 0.45229461789131165, | |
| "learning_rate": 0.0002687280139778683, | |
| "loss": 3.1651, | |
| "step": 94900 | |
| }, | |
| { | |
| "epoch": 27.63402421984164, | |
| "grad_norm": 0.43945810198783875, | |
| "learning_rate": 0.00026855329062317997, | |
| "loss": 3.1684, | |
| "step": 94950 | |
| }, | |
| { | |
| "epoch": 27.648579413134605, | |
| "grad_norm": 0.46427789330482483, | |
| "learning_rate": 0.0002683785672684915, | |
| "loss": 3.1656, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.648579413134605, | |
| "eval_accuracy": 0.375150332114536, | |
| "eval_loss": 3.535543918609619, | |
| "eval_runtime": 82.842, | |
| "eval_samples_per_second": 200.997, | |
| "eval_steps_per_second": 12.566, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.663134606427572, | |
| "grad_norm": 0.44693025946617126, | |
| "learning_rate": 0.0002682038439138031, | |
| "loss": 3.1734, | |
| "step": 95050 | |
| }, | |
| { | |
| "epoch": 27.67768979972054, | |
| "grad_norm": 0.4252467453479767, | |
| "learning_rate": 0.0002680291205591147, | |
| "loss": 3.1645, | |
| "step": 95100 | |
| }, | |
| { | |
| "epoch": 27.692244993013507, | |
| "grad_norm": 0.42376407980918884, | |
| "learning_rate": 0.0002678543972044263, | |
| "loss": 3.1732, | |
| "step": 95150 | |
| }, | |
| { | |
| "epoch": 27.706800186306474, | |
| "grad_norm": 0.45157590508461, | |
| "learning_rate": 0.0002676796738497379, | |
| "loss": 3.1783, | |
| "step": 95200 | |
| }, | |
| { | |
| "epoch": 27.72135537959944, | |
| "grad_norm": 0.4422607421875, | |
| "learning_rate": 0.0002675049504950495, | |
| "loss": 3.1774, | |
| "step": 95250 | |
| }, | |
| { | |
| "epoch": 27.735910572892408, | |
| "grad_norm": 0.4435870349407196, | |
| "learning_rate": 0.0002673302271403611, | |
| "loss": 3.1696, | |
| "step": 95300 | |
| }, | |
| { | |
| "epoch": 27.750465766185375, | |
| "grad_norm": 0.4601503312587738, | |
| "learning_rate": 0.00026715550378567267, | |
| "loss": 3.1848, | |
| "step": 95350 | |
| }, | |
| { | |
| "epoch": 27.765020959478342, | |
| "grad_norm": 0.448876291513443, | |
| "learning_rate": 0.00026698078043098426, | |
| "loss": 3.165, | |
| "step": 95400 | |
| }, | |
| { | |
| "epoch": 27.77957615277131, | |
| "grad_norm": 0.45465195178985596, | |
| "learning_rate": 0.00026680605707629586, | |
| "loss": 3.1659, | |
| "step": 95450 | |
| }, | |
| { | |
| "epoch": 27.794131346064276, | |
| "grad_norm": 0.4470723271369934, | |
| "learning_rate": 0.00026663133372160745, | |
| "loss": 3.1852, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 27.808686539357243, | |
| "grad_norm": 0.4297708570957184, | |
| "learning_rate": 0.00026645661036691905, | |
| "loss": 3.177, | |
| "step": 95550 | |
| }, | |
| { | |
| "epoch": 27.82324173265021, | |
| "grad_norm": 0.4115425944328308, | |
| "learning_rate": 0.0002662818870122306, | |
| "loss": 3.1706, | |
| "step": 95600 | |
| }, | |
| { | |
| "epoch": 27.837796925943177, | |
| "grad_norm": 0.47041401267051697, | |
| "learning_rate": 0.00026610716365754224, | |
| "loss": 3.1679, | |
| "step": 95650 | |
| }, | |
| { | |
| "epoch": 27.852352119236144, | |
| "grad_norm": 0.4212128818035126, | |
| "learning_rate": 0.0002659324403028538, | |
| "loss": 3.1826, | |
| "step": 95700 | |
| }, | |
| { | |
| "epoch": 27.86690731252911, | |
| "grad_norm": 0.4273899495601654, | |
| "learning_rate": 0.00026575771694816537, | |
| "loss": 3.1849, | |
| "step": 95750 | |
| }, | |
| { | |
| "epoch": 27.88146250582208, | |
| "grad_norm": 0.44989147782325745, | |
| "learning_rate": 0.00026558299359347696, | |
| "loss": 3.1808, | |
| "step": 95800 | |
| }, | |
| { | |
| "epoch": 27.896017699115045, | |
| "grad_norm": 0.4507368206977844, | |
| "learning_rate": 0.00026540827023878856, | |
| "loss": 3.1862, | |
| "step": 95850 | |
| }, | |
| { | |
| "epoch": 27.910572892408013, | |
| "grad_norm": 0.4297460913658142, | |
| "learning_rate": 0.00026523354688410015, | |
| "loss": 3.1798, | |
| "step": 95900 | |
| }, | |
| { | |
| "epoch": 27.92512808570098, | |
| "grad_norm": 0.4367275834083557, | |
| "learning_rate": 0.00026505882352941175, | |
| "loss": 3.1977, | |
| "step": 95950 | |
| }, | |
| { | |
| "epoch": 27.939683278993947, | |
| "grad_norm": 0.4250665307044983, | |
| "learning_rate": 0.00026488410017472334, | |
| "loss": 3.1849, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.939683278993947, | |
| "eval_accuracy": 0.37556320553845074, | |
| "eval_loss": 3.532724380493164, | |
| "eval_runtime": 82.8, | |
| "eval_samples_per_second": 201.099, | |
| "eval_steps_per_second": 12.572, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.95423847228691, | |
| "grad_norm": 0.44175103306770325, | |
| "learning_rate": 0.00026470937682003494, | |
| "loss": 3.1831, | |
| "step": 96050 | |
| }, | |
| { | |
| "epoch": 27.968793665579877, | |
| "grad_norm": 0.44950711727142334, | |
| "learning_rate": 0.00026453465346534653, | |
| "loss": 3.1848, | |
| "step": 96100 | |
| }, | |
| { | |
| "epoch": 27.983348858872844, | |
| "grad_norm": 0.4530915915966034, | |
| "learning_rate": 0.00026435993011065807, | |
| "loss": 3.1862, | |
| "step": 96150 | |
| }, | |
| { | |
| "epoch": 27.99790405216581, | |
| "grad_norm": 0.4433324933052063, | |
| "learning_rate": 0.0002641852067559697, | |
| "loss": 3.1815, | |
| "step": 96200 | |
| }, | |
| { | |
| "epoch": 28.012226362366093, | |
| "grad_norm": 0.45721039175987244, | |
| "learning_rate": 0.00026401048340128126, | |
| "loss": 3.0827, | |
| "step": 96250 | |
| }, | |
| { | |
| "epoch": 28.02678155565906, | |
| "grad_norm": 0.4741598665714264, | |
| "learning_rate": 0.00026383576004659285, | |
| "loss": 3.0933, | |
| "step": 96300 | |
| }, | |
| { | |
| "epoch": 28.041336748952027, | |
| "grad_norm": 0.4262336194515228, | |
| "learning_rate": 0.00026366103669190445, | |
| "loss": 3.0945, | |
| "step": 96350 | |
| }, | |
| { | |
| "epoch": 28.055891942244994, | |
| "grad_norm": 0.4419623613357544, | |
| "learning_rate": 0.00026348631333721604, | |
| "loss": 3.1116, | |
| "step": 96400 | |
| }, | |
| { | |
| "epoch": 28.07044713553796, | |
| "grad_norm": 0.4444783627986908, | |
| "learning_rate": 0.00026331158998252764, | |
| "loss": 3.0991, | |
| "step": 96450 | |
| }, | |
| { | |
| "epoch": 28.085002328830928, | |
| "grad_norm": 0.4661199450492859, | |
| "learning_rate": 0.00026313686662783923, | |
| "loss": 3.1016, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 28.099557522123895, | |
| "grad_norm": 0.4601952135562897, | |
| "learning_rate": 0.0002629621432731508, | |
| "loss": 3.0977, | |
| "step": 96550 | |
| }, | |
| { | |
| "epoch": 28.114112715416862, | |
| "grad_norm": 0.4516938328742981, | |
| "learning_rate": 0.0002627874199184624, | |
| "loss": 3.1138, | |
| "step": 96600 | |
| }, | |
| { | |
| "epoch": 28.12866790870983, | |
| "grad_norm": 0.45379379391670227, | |
| "learning_rate": 0.000262612696563774, | |
| "loss": 3.1033, | |
| "step": 96650 | |
| }, | |
| { | |
| "epoch": 28.143223102002796, | |
| "grad_norm": 0.4646652638912201, | |
| "learning_rate": 0.0002624379732090856, | |
| "loss": 3.1175, | |
| "step": 96700 | |
| }, | |
| { | |
| "epoch": 28.15777829529576, | |
| "grad_norm": 0.4311061203479767, | |
| "learning_rate": 0.0002622632498543972, | |
| "loss": 3.1095, | |
| "step": 96750 | |
| }, | |
| { | |
| "epoch": 28.172333488588727, | |
| "grad_norm": 0.47682809829711914, | |
| "learning_rate": 0.0002620885264997088, | |
| "loss": 3.1177, | |
| "step": 96800 | |
| }, | |
| { | |
| "epoch": 28.186888681881694, | |
| "grad_norm": 0.4490298628807068, | |
| "learning_rate": 0.00026191380314502034, | |
| "loss": 3.1265, | |
| "step": 96850 | |
| }, | |
| { | |
| "epoch": 28.20144387517466, | |
| "grad_norm": 0.4710911512374878, | |
| "learning_rate": 0.000261739079790332, | |
| "loss": 3.0995, | |
| "step": 96900 | |
| }, | |
| { | |
| "epoch": 28.215999068467628, | |
| "grad_norm": 0.4507644474506378, | |
| "learning_rate": 0.0002615643564356435, | |
| "loss": 3.1267, | |
| "step": 96950 | |
| }, | |
| { | |
| "epoch": 28.230554261760595, | |
| "grad_norm": 0.4267343580722809, | |
| "learning_rate": 0.0002613896330809551, | |
| "loss": 3.1252, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.230554261760595, | |
| "eval_accuracy": 0.3746350924076068, | |
| "eval_loss": 3.551305055618286, | |
| "eval_runtime": 82.8518, | |
| "eval_samples_per_second": 200.973, | |
| "eval_steps_per_second": 12.565, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.245109455053562, | |
| "grad_norm": 0.4779544174671173, | |
| "learning_rate": 0.0002612149097262667, | |
| "loss": 3.1315, | |
| "step": 97050 | |
| }, | |
| { | |
| "epoch": 28.25966464834653, | |
| "grad_norm": 0.44766417145729065, | |
| "learning_rate": 0.0002610401863715783, | |
| "loss": 3.1186, | |
| "step": 97100 | |
| }, | |
| { | |
| "epoch": 28.274219841639496, | |
| "grad_norm": 0.42847368121147156, | |
| "learning_rate": 0.0002608654630168899, | |
| "loss": 3.1256, | |
| "step": 97150 | |
| }, | |
| { | |
| "epoch": 28.288775034932463, | |
| "grad_norm": 0.46162867546081543, | |
| "learning_rate": 0.0002606907396622015, | |
| "loss": 3.1368, | |
| "step": 97200 | |
| }, | |
| { | |
| "epoch": 28.30333022822543, | |
| "grad_norm": 0.5105154514312744, | |
| "learning_rate": 0.0002605160163075131, | |
| "loss": 3.1248, | |
| "step": 97250 | |
| }, | |
| { | |
| "epoch": 28.317885421518397, | |
| "grad_norm": 0.47541460394859314, | |
| "learning_rate": 0.0002603412929528247, | |
| "loss": 3.1395, | |
| "step": 97300 | |
| }, | |
| { | |
| "epoch": 28.332440614811365, | |
| "grad_norm": 0.4751850366592407, | |
| "learning_rate": 0.0002601665695981363, | |
| "loss": 3.1393, | |
| "step": 97350 | |
| }, | |
| { | |
| "epoch": 28.34699580810433, | |
| "grad_norm": 0.467650830745697, | |
| "learning_rate": 0.0002599918462434478, | |
| "loss": 3.1417, | |
| "step": 97400 | |
| }, | |
| { | |
| "epoch": 28.3615510013973, | |
| "grad_norm": 0.4525465965270996, | |
| "learning_rate": 0.00025981712288875947, | |
| "loss": 3.136, | |
| "step": 97450 | |
| }, | |
| { | |
| "epoch": 28.376106194690266, | |
| "grad_norm": 0.4389481544494629, | |
| "learning_rate": 0.000259642399534071, | |
| "loss": 3.1493, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 28.390661387983233, | |
| "grad_norm": 0.44302141666412354, | |
| "learning_rate": 0.0002594676761793826, | |
| "loss": 3.144, | |
| "step": 97550 | |
| }, | |
| { | |
| "epoch": 28.4052165812762, | |
| "grad_norm": 0.44878682494163513, | |
| "learning_rate": 0.0002592929528246942, | |
| "loss": 3.1477, | |
| "step": 97600 | |
| }, | |
| { | |
| "epoch": 28.419771774569167, | |
| "grad_norm": 0.47434788942337036, | |
| "learning_rate": 0.0002591182294700058, | |
| "loss": 3.1445, | |
| "step": 97650 | |
| }, | |
| { | |
| "epoch": 28.434326967862134, | |
| "grad_norm": 0.4481390416622162, | |
| "learning_rate": 0.0002589435061153174, | |
| "loss": 3.1506, | |
| "step": 97700 | |
| }, | |
| { | |
| "epoch": 28.4488821611551, | |
| "grad_norm": 0.440329372882843, | |
| "learning_rate": 0.000258768782760629, | |
| "loss": 3.1452, | |
| "step": 97750 | |
| }, | |
| { | |
| "epoch": 28.463437354448068, | |
| "grad_norm": 0.4370376467704773, | |
| "learning_rate": 0.0002585940594059406, | |
| "loss": 3.1499, | |
| "step": 97800 | |
| }, | |
| { | |
| "epoch": 28.477992547741035, | |
| "grad_norm": 0.47814446687698364, | |
| "learning_rate": 0.00025841933605125217, | |
| "loss": 3.153, | |
| "step": 97850 | |
| }, | |
| { | |
| "epoch": 28.492547741034002, | |
| "grad_norm": 0.4670361876487732, | |
| "learning_rate": 0.00025824461269656376, | |
| "loss": 3.1589, | |
| "step": 97900 | |
| }, | |
| { | |
| "epoch": 28.50710293432697, | |
| "grad_norm": 0.4508378803730011, | |
| "learning_rate": 0.00025806988934187536, | |
| "loss": 3.1466, | |
| "step": 97950 | |
| }, | |
| { | |
| "epoch": 28.521658127619936, | |
| "grad_norm": 0.44208523631095886, | |
| "learning_rate": 0.00025789516598718695, | |
| "loss": 3.1448, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.521658127619936, | |
| "eval_accuracy": 0.3751678436830425, | |
| "eval_loss": 3.5449936389923096, | |
| "eval_runtime": 83.0691, | |
| "eval_samples_per_second": 200.447, | |
| "eval_steps_per_second": 12.532, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.536213320912903, | |
| "grad_norm": 0.47016555070877075, | |
| "learning_rate": 0.00025772044263249854, | |
| "loss": 3.1604, | |
| "step": 98050 | |
| }, | |
| { | |
| "epoch": 28.55076851420587, | |
| "grad_norm": 0.43794775009155273, | |
| "learning_rate": 0.0002575457192778101, | |
| "loss": 3.1636, | |
| "step": 98100 | |
| }, | |
| { | |
| "epoch": 28.565323707498834, | |
| "grad_norm": 0.45240283012390137, | |
| "learning_rate": 0.00025737099592312173, | |
| "loss": 3.1415, | |
| "step": 98150 | |
| }, | |
| { | |
| "epoch": 28.5798789007918, | |
| "grad_norm": 0.45302581787109375, | |
| "learning_rate": 0.0002571962725684333, | |
| "loss": 3.1573, | |
| "step": 98200 | |
| }, | |
| { | |
| "epoch": 28.594434094084768, | |
| "grad_norm": 0.4644310772418976, | |
| "learning_rate": 0.00025702154921374487, | |
| "loss": 3.1587, | |
| "step": 98250 | |
| }, | |
| { | |
| "epoch": 28.608989287377735, | |
| "grad_norm": 0.4451628625392914, | |
| "learning_rate": 0.00025684682585905646, | |
| "loss": 3.1584, | |
| "step": 98300 | |
| }, | |
| { | |
| "epoch": 28.623544480670702, | |
| "grad_norm": 0.4455444812774658, | |
| "learning_rate": 0.00025667210250436806, | |
| "loss": 3.1712, | |
| "step": 98350 | |
| }, | |
| { | |
| "epoch": 28.63809967396367, | |
| "grad_norm": 0.49326446652412415, | |
| "learning_rate": 0.00025649737914967965, | |
| "loss": 3.1658, | |
| "step": 98400 | |
| }, | |
| { | |
| "epoch": 28.652654867256636, | |
| "grad_norm": 0.47082576155662537, | |
| "learning_rate": 0.00025632265579499124, | |
| "loss": 3.1727, | |
| "step": 98450 | |
| }, | |
| { | |
| "epoch": 28.667210060549603, | |
| "grad_norm": 0.4508187770843506, | |
| "learning_rate": 0.00025614793244030284, | |
| "loss": 3.1572, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 28.68176525384257, | |
| "grad_norm": 0.44190704822540283, | |
| "learning_rate": 0.0002559732090856144, | |
| "loss": 3.1571, | |
| "step": 98550 | |
| }, | |
| { | |
| "epoch": 28.696320447135538, | |
| "grad_norm": 0.47588956356048584, | |
| "learning_rate": 0.00025579848573092603, | |
| "loss": 3.1574, | |
| "step": 98600 | |
| }, | |
| { | |
| "epoch": 28.710875640428505, | |
| "grad_norm": 0.444128155708313, | |
| "learning_rate": 0.0002556237623762376, | |
| "loss": 3.1773, | |
| "step": 98650 | |
| }, | |
| { | |
| "epoch": 28.72543083372147, | |
| "grad_norm": 0.46229106187820435, | |
| "learning_rate": 0.0002554490390215492, | |
| "loss": 3.1738, | |
| "step": 98700 | |
| }, | |
| { | |
| "epoch": 28.73998602701444, | |
| "grad_norm": 0.4813465476036072, | |
| "learning_rate": 0.0002552743156668608, | |
| "loss": 3.1774, | |
| "step": 98750 | |
| }, | |
| { | |
| "epoch": 28.754541220307406, | |
| "grad_norm": 0.4433896541595459, | |
| "learning_rate": 0.00025509959231217235, | |
| "loss": 3.1692, | |
| "step": 98800 | |
| }, | |
| { | |
| "epoch": 28.769096413600373, | |
| "grad_norm": 0.46264949440956116, | |
| "learning_rate": 0.000254924868957484, | |
| "loss": 3.165, | |
| "step": 98850 | |
| }, | |
| { | |
| "epoch": 28.78365160689334, | |
| "grad_norm": 0.4496705234050751, | |
| "learning_rate": 0.00025475014560279554, | |
| "loss": 3.1802, | |
| "step": 98900 | |
| }, | |
| { | |
| "epoch": 28.798206800186307, | |
| "grad_norm": 0.44545769691467285, | |
| "learning_rate": 0.00025457542224810713, | |
| "loss": 3.1644, | |
| "step": 98950 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "grad_norm": 0.4406527578830719, | |
| "learning_rate": 0.00025440069889341873, | |
| "loss": 3.1654, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "eval_accuracy": 0.37546001656429845, | |
| "eval_loss": 3.535421133041382, | |
| "eval_runtime": 82.9003, | |
| "eval_samples_per_second": 200.856, | |
| "eval_steps_per_second": 12.557, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "step": 99000, | |
| "total_flos": 2.068898812526592e+18, | |
| "train_loss": 3.392319653597745, | |
| "train_runtime": 73667.1201, | |
| "train_samples_per_second": 186.519, | |
| "train_steps_per_second": 2.332 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 171800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 20, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 11 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.068898812526592e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |