Invalid JSON: Unexpected token 'N', ..."ad_norm": NaN,
"... is not valid JSON
| { | |
| "best_global_step": 79000, | |
| "best_metric": 3.5298874378204346, | |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/swap_require_to_drop_2128/checkpoint-40000", | |
| "epoch": 32.01426408942711, | |
| "eval_steps": 1000, | |
| "global_step": 110000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014555193292966931, | |
| "grad_norm": 1.664777398109436, | |
| "learning_rate": 0.000294, | |
| "loss": 8.4459, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029110386585933862, | |
| "grad_norm": 0.8744126558303833, | |
| "learning_rate": 0.0005939999999999999, | |
| "loss": 6.7213, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04366557987890079, | |
| "grad_norm": 0.514337956905365, | |
| "learning_rate": 0.0005998287711124053, | |
| "loss": 6.3296, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.058220773171867725, | |
| "grad_norm": 0.584080696105957, | |
| "learning_rate": 0.000599654047757717, | |
| "loss": 6.151, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07277596646483465, | |
| "grad_norm": 0.45835795998573303, | |
| "learning_rate": 0.0005994793244030285, | |
| "loss": 5.9916, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.08733115975780158, | |
| "grad_norm": 0.4237825870513916, | |
| "learning_rate": 0.00059930460104834, | |
| "loss": 5.8717, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.10188635305076851, | |
| "grad_norm": 0.410149484872818, | |
| "learning_rate": 0.0005991298776936517, | |
| "loss": 5.7453, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.11644154634373545, | |
| "grad_norm": 0.5075629949569702, | |
| "learning_rate": 0.0005989551543389632, | |
| "loss": 5.6456, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1309967396367024, | |
| "grad_norm": 0.45372700691223145, | |
| "learning_rate": 0.0005987804309842748, | |
| "loss": 5.5133, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1455519329296693, | |
| "grad_norm": 0.45813828706741333, | |
| "learning_rate": 0.0005986057076295864, | |
| "loss": 5.4275, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16010712622263623, | |
| "grad_norm": 0.42607325315475464, | |
| "learning_rate": 0.0005984309842748981, | |
| "loss": 5.3298, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.17466231951560315, | |
| "grad_norm": 0.5173180103302002, | |
| "learning_rate": 0.0005982562609202096, | |
| "loss": 5.2464, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1892175128085701, | |
| "grad_norm": 0.43883052468299866, | |
| "learning_rate": 0.0005980815375655212, | |
| "loss": 5.1861, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.20377270610153703, | |
| "grad_norm": 0.4173907935619354, | |
| "learning_rate": 0.0005979068142108328, | |
| "loss": 5.1403, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.21832789939450395, | |
| "grad_norm": 0.4811343550682068, | |
| "learning_rate": 0.0005977320908561445, | |
| "loss": 5.0862, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2328830926874709, | |
| "grad_norm": 0.4635143280029297, | |
| "learning_rate": 0.000597557367501456, | |
| "loss": 5.0425, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.24743828598043782, | |
| "grad_norm": 0.37773096561431885, | |
| "learning_rate": 0.0005973826441467675, | |
| "loss": 4.9656, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2619934792734048, | |
| "grad_norm": 0.4224317669868469, | |
| "learning_rate": 0.0005972079207920792, | |
| "loss": 4.9229, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.27654867256637167, | |
| "grad_norm": 0.4564674496650696, | |
| "learning_rate": 0.0005970331974373907, | |
| "loss": 4.8711, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "grad_norm": 0.44638097286224365, | |
| "learning_rate": 0.0005968584740827023, | |
| "loss": 4.8313, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "eval_accuracy": 0.25401106002460316, | |
| "eval_loss": 4.755082130432129, | |
| "eval_runtime": 180.1011, | |
| "eval_samples_per_second": 92.454, | |
| "eval_steps_per_second": 5.78, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.30565905915230557, | |
| "grad_norm": 0.5874536037445068, | |
| "learning_rate": 0.0005966837507280139, | |
| "loss": 4.7718, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.32021425244527246, | |
| "grad_norm": 0.4394214451313019, | |
| "learning_rate": 0.0005965090273733256, | |
| "loss": 4.7342, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3347694457382394, | |
| "grad_norm": 0.44372034072875977, | |
| "learning_rate": 0.0005963343040186371, | |
| "loss": 4.7015, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3493246390312063, | |
| "grad_norm": 0.4285391867160797, | |
| "learning_rate": 0.0005961595806639486, | |
| "loss": 4.6587, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.36387983232417326, | |
| "grad_norm": 0.4155198037624359, | |
| "learning_rate": 0.0005959848573092603, | |
| "loss": 4.6256, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.3784350256171402, | |
| "grad_norm": 0.4572656452655792, | |
| "learning_rate": 0.0005958101339545718, | |
| "loss": 4.6061, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3929902189101071, | |
| "grad_norm": 0.4733801484107971, | |
| "learning_rate": 0.0005956354105998835, | |
| "loss": 4.5809, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.40754541220307405, | |
| "grad_norm": 0.41797181963920593, | |
| "learning_rate": 0.000595460687245195, | |
| "loss": 4.5521, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.422100605496041, | |
| "grad_norm": 0.461195707321167, | |
| "learning_rate": 0.0005952859638905067, | |
| "loss": 4.5162, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.4366557987890079, | |
| "grad_norm": 0.43522897362709045, | |
| "learning_rate": 0.0005951112405358182, | |
| "loss": 4.4957, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.45121099208197485, | |
| "grad_norm": 0.43900424242019653, | |
| "learning_rate": 0.0005949365171811299, | |
| "loss": 4.4926, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4657661853749418, | |
| "grad_norm": 0.42797166109085083, | |
| "learning_rate": 0.0005947617938264414, | |
| "loss": 4.4513, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.4803213786679087, | |
| "grad_norm": 0.4188319444656372, | |
| "learning_rate": 0.000594587070471753, | |
| "loss": 4.4385, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.49487657196087564, | |
| "grad_norm": 0.4120756685733795, | |
| "learning_rate": 0.0005944123471170646, | |
| "loss": 4.4206, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5094317652538426, | |
| "grad_norm": 0.47579753398895264, | |
| "learning_rate": 0.0005942376237623762, | |
| "loss": 4.4176, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5239869585468095, | |
| "grad_norm": 0.43895846605300903, | |
| "learning_rate": 0.0005940629004076878, | |
| "loss": 4.3887, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5385421518397764, | |
| "grad_norm": 0.39533790946006775, | |
| "learning_rate": 0.0005938881770529993, | |
| "loss": 4.38, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5530973451327433, | |
| "grad_norm": 0.377754271030426, | |
| "learning_rate": 0.000593713453698311, | |
| "loss": 4.3735, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.5676525384257103, | |
| "grad_norm": 0.39021018147468567, | |
| "learning_rate": 0.0005935387303436226, | |
| "loss": 4.3428, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "grad_norm": 0.3872169554233551, | |
| "learning_rate": 0.0005933640069889342, | |
| "loss": 4.327, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "eval_accuracy": 0.2996545519912005, | |
| "eval_loss": 4.280758380889893, | |
| "eval_runtime": 179.3832, | |
| "eval_samples_per_second": 92.824, | |
| "eval_steps_per_second": 5.803, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5967629250116442, | |
| "grad_norm": 0.42003875970840454, | |
| "learning_rate": 0.0005931892836342457, | |
| "loss": 4.3191, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.6113181183046111, | |
| "grad_norm": 0.38116443157196045, | |
| "learning_rate": 0.0005930145602795573, | |
| "loss": 4.3089, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.625873311597578, | |
| "grad_norm": 0.3790656328201294, | |
| "learning_rate": 0.000592839836924869, | |
| "loss": 4.2951, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6404285048905449, | |
| "grad_norm": 0.40056294202804565, | |
| "learning_rate": 0.0005926651135701805, | |
| "loss": 4.2845, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.6549836981835119, | |
| "grad_norm": 0.3878796398639679, | |
| "learning_rate": 0.0005924903902154921, | |
| "loss": 4.2657, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6695388914764788, | |
| "grad_norm": 0.388304740190506, | |
| "learning_rate": 0.0005923156668608037, | |
| "loss": 4.2631, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.6840940847694458, | |
| "grad_norm": 0.41076818108558655, | |
| "learning_rate": 0.0005921409435061153, | |
| "loss": 4.2714, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6986492780624126, | |
| "grad_norm": 0.3966241776943207, | |
| "learning_rate": 0.0005919662201514268, | |
| "loss": 4.2525, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7132044713553796, | |
| "grad_norm": 0.39895981550216675, | |
| "learning_rate": 0.0005917914967967384, | |
| "loss": 4.2249, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.7277596646483465, | |
| "grad_norm": 0.37347468733787537, | |
| "learning_rate": 0.0005916167734420501, | |
| "loss": 4.2305, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7423148579413135, | |
| "grad_norm": 0.36039796471595764, | |
| "learning_rate": 0.0005914420500873616, | |
| "loss": 4.2042, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7568700512342804, | |
| "grad_norm": 0.37558725476264954, | |
| "learning_rate": 0.0005912673267326732, | |
| "loss": 4.208, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.7714252445272474, | |
| "grad_norm": 0.38653722405433655, | |
| "learning_rate": 0.0005910926033779848, | |
| "loss": 4.1975, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.7859804378202142, | |
| "grad_norm": 0.37924301624298096, | |
| "learning_rate": 0.0005909178800232964, | |
| "loss": 4.1807, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8005356311131812, | |
| "grad_norm": 0.378798246383667, | |
| "learning_rate": 0.000590743156668608, | |
| "loss": 4.1775, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.8150908244061481, | |
| "grad_norm": 0.35967278480529785, | |
| "learning_rate": 0.0005905684333139196, | |
| "loss": 4.1639, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8296460176991151, | |
| "grad_norm": 0.3435153663158417, | |
| "learning_rate": 0.0005903937099592312, | |
| "loss": 4.1748, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.844201210992082, | |
| "grad_norm": 0.3642827272415161, | |
| "learning_rate": 0.0005902189866045427, | |
| "loss": 4.1575, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.858756404285049, | |
| "grad_norm": 0.34654879570007324, | |
| "learning_rate": 0.0005900442632498543, | |
| "loss": 4.1613, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "grad_norm": 0.37901315093040466, | |
| "learning_rate": 0.0005898695398951659, | |
| "loss": 4.143, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "eval_accuracy": 0.3158166719769421, | |
| "eval_loss": 4.093630313873291, | |
| "eval_runtime": 179.4838, | |
| "eval_samples_per_second": 92.772, | |
| "eval_steps_per_second": 5.8, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8878667908709827, | |
| "grad_norm": 0.35573050379753113, | |
| "learning_rate": 0.0005896948165404776, | |
| "loss": 4.1588, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.9024219841639497, | |
| "grad_norm": 0.37339988350868225, | |
| "learning_rate": 0.0005895200931857891, | |
| "loss": 4.1301, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9169771774569166, | |
| "grad_norm": 0.36611881852149963, | |
| "learning_rate": 0.0005893453698311007, | |
| "loss": 4.1326, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9315323707498836, | |
| "grad_norm": 0.3458259403705597, | |
| "learning_rate": 0.0005891706464764123, | |
| "loss": 4.1098, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.9460875640428504, | |
| "grad_norm": 0.34497207403182983, | |
| "learning_rate": 0.0005889959231217238, | |
| "loss": 4.11, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.9606427573358174, | |
| "grad_norm": 0.3454968333244324, | |
| "learning_rate": 0.0005888211997670355, | |
| "loss": 4.0968, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.9751979506287843, | |
| "grad_norm": 0.3598727285861969, | |
| "learning_rate": 0.000588646476412347, | |
| "loss": 4.0848, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.9897531439217513, | |
| "grad_norm": 0.37226811051368713, | |
| "learning_rate": 0.0005884717530576587, | |
| "loss": 4.1, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.0040754541220307, | |
| "grad_norm": 0.36724382638931274, | |
| "learning_rate": 0.0005882970297029702, | |
| "loss": 4.0639, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.0186306474149978, | |
| "grad_norm": 0.3470579981803894, | |
| "learning_rate": 0.0005881223063482818, | |
| "loss": 4.0231, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.0331858407079646, | |
| "grad_norm": 0.37876060605049133, | |
| "learning_rate": 0.0005879475829935934, | |
| "loss": 4.015, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.0477410340009314, | |
| "grad_norm": 0.3453887104988098, | |
| "learning_rate": 0.0005877728596389051, | |
| "loss": 4.0009, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.0622962272938985, | |
| "grad_norm": 0.35795602202415466, | |
| "learning_rate": 0.0005875981362842166, | |
| "loss": 4.0177, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.0768514205868653, | |
| "grad_norm": 0.3534950017929077, | |
| "learning_rate": 0.0005874234129295281, | |
| "loss": 4.0121, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.0914066138798324, | |
| "grad_norm": 0.3307143747806549, | |
| "learning_rate": 0.0005872486895748398, | |
| "loss": 4.0045, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.1059618071727992, | |
| "grad_norm": 0.34968075156211853, | |
| "learning_rate": 0.0005870739662201513, | |
| "loss": 3.9982, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.120517000465766, | |
| "grad_norm": 0.3467692732810974, | |
| "learning_rate": 0.000586899242865463, | |
| "loss": 3.9971, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.1350721937587331, | |
| "grad_norm": 0.34137192368507385, | |
| "learning_rate": 0.0005867245195107746, | |
| "loss": 4.0123, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.1496273870517, | |
| "grad_norm": 0.3370218276977539, | |
| "learning_rate": 0.0005865497961560862, | |
| "loss": 3.9948, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "grad_norm": 0.35244500637054443, | |
| "learning_rate": 0.0005863750728013977, | |
| "loss": 3.9875, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "eval_accuracy": 0.32581989104983733, | |
| "eval_loss": 3.9877822399139404, | |
| "eval_runtime": 179.4823, | |
| "eval_samples_per_second": 92.772, | |
| "eval_steps_per_second": 5.8, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.1787377736376339, | |
| "grad_norm": 0.34883835911750793, | |
| "learning_rate": 0.0005862003494467094, | |
| "loss": 3.991, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.193292966930601, | |
| "grad_norm": 0.33956724405288696, | |
| "learning_rate": 0.0005860256260920209, | |
| "loss": 3.971, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.2078481602235678, | |
| "grad_norm": 0.3573682904243469, | |
| "learning_rate": 0.0005858509027373325, | |
| "loss": 3.97, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 1.2224033535165346, | |
| "grad_norm": 0.340936541557312, | |
| "learning_rate": 0.0005856761793826441, | |
| "loss": 3.9766, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.2369585468095017, | |
| "grad_norm": 0.34760385751724243, | |
| "learning_rate": 0.0005855014560279557, | |
| "loss": 3.9639, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 1.2515137401024685, | |
| "grad_norm": 0.3335438668727875, | |
| "learning_rate": 0.0005853267326732673, | |
| "loss": 3.9661, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.2660689333954354, | |
| "grad_norm": 0.3376685380935669, | |
| "learning_rate": 0.0005851520093185788, | |
| "loss": 3.965, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 1.2806241266884024, | |
| "grad_norm": 0.3536045253276825, | |
| "learning_rate": 0.0005849772859638905, | |
| "loss": 3.9526, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.2951793199813695, | |
| "grad_norm": 0.3248458802700043, | |
| "learning_rate": 0.0005848025626092021, | |
| "loss": 3.9497, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 1.3097345132743363, | |
| "grad_norm": 0.33945971727371216, | |
| "learning_rate": 0.0005846278392545136, | |
| "loss": 3.9642, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.3242897065673032, | |
| "grad_norm": 0.320038378238678, | |
| "learning_rate": 0.0005844531158998252, | |
| "loss": 3.9586, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.3388448998602702, | |
| "grad_norm": 0.3388155996799469, | |
| "learning_rate": 0.0005842783925451368, | |
| "loss": 3.9434, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.353400093153237, | |
| "grad_norm": 0.34264621138572693, | |
| "learning_rate": 0.0005841036691904484, | |
| "loss": 3.9458, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.367955286446204, | |
| "grad_norm": 0.3136458992958069, | |
| "learning_rate": 0.00058392894583576, | |
| "loss": 3.9506, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.382510479739171, | |
| "grad_norm": 0.34289732575416565, | |
| "learning_rate": 0.0005837542224810716, | |
| "loss": 3.9497, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 1.3970656730321378, | |
| "grad_norm": 0.36164990067481995, | |
| "learning_rate": 0.0005835794991263832, | |
| "loss": 3.9397, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.4116208663251049, | |
| "grad_norm": 0.34485942125320435, | |
| "learning_rate": 0.0005834047757716948, | |
| "loss": 3.9358, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 1.4261760596180717, | |
| "grad_norm": 0.3529173731803894, | |
| "learning_rate": 0.0005832300524170063, | |
| "loss": 3.9443, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.4407312529110388, | |
| "grad_norm": 0.32209211587905884, | |
| "learning_rate": 0.0005830553290623179, | |
| "loss": 3.9202, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "grad_norm": 0.3291449546813965, | |
| "learning_rate": 0.0005828806057076296, | |
| "loss": 3.9281, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "eval_accuracy": 0.3324244555048086, | |
| "eval_loss": 3.9115827083587646, | |
| "eval_runtime": 179.3171, | |
| "eval_samples_per_second": 92.858, | |
| "eval_steps_per_second": 5.805, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4698416394969724, | |
| "grad_norm": 0.32161611318588257, | |
| "learning_rate": 0.0005827058823529411, | |
| "loss": 3.92, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 1.4843968327899395, | |
| "grad_norm": 0.321431428194046, | |
| "learning_rate": 0.0005825311589982527, | |
| "loss": 3.914, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.4989520260829063, | |
| "grad_norm": 0.33909738063812256, | |
| "learning_rate": 0.0005823564356435643, | |
| "loss": 3.9124, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 1.5135072193758732, | |
| "grad_norm": 0.3282327353954315, | |
| "learning_rate": 0.0005821817122888759, | |
| "loss": 3.9224, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.5280624126688402, | |
| "grad_norm": 0.32862597703933716, | |
| "learning_rate": 0.0005820069889341875, | |
| "loss": 3.9115, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 1.5426176059618073, | |
| "grad_norm": 0.33926936984062195, | |
| "learning_rate": 0.000581832265579499, | |
| "loss": 3.9111, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.5571727992547741, | |
| "grad_norm": 0.31386932730674744, | |
| "learning_rate": 0.0005816575422248107, | |
| "loss": 3.9139, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 1.571727992547741, | |
| "grad_norm": 0.32861968874931335, | |
| "learning_rate": 0.0005814828188701222, | |
| "loss": 3.904, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.586283185840708, | |
| "grad_norm": 0.3435116112232208, | |
| "learning_rate": 0.0005813080955154338, | |
| "loss": 3.9043, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.6008383791336749, | |
| "grad_norm": 0.3164297640323639, | |
| "learning_rate": 0.0005811333721607454, | |
| "loss": 3.9041, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.6153935724266417, | |
| "grad_norm": 0.325747549533844, | |
| "learning_rate": 0.0005809586488060571, | |
| "loss": 3.8981, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.6299487657196088, | |
| "grad_norm": 0.3419976532459259, | |
| "learning_rate": 0.0005807839254513686, | |
| "loss": 3.8924, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.6445039590125758, | |
| "grad_norm": 0.3399415612220764, | |
| "learning_rate": 0.0005806092020966802, | |
| "loss": 3.8939, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.6590591523055425, | |
| "grad_norm": 0.31317266821861267, | |
| "learning_rate": 0.0005804344787419918, | |
| "loss": 3.8892, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.6736143455985095, | |
| "grad_norm": 0.33786946535110474, | |
| "learning_rate": 0.0005802597553873033, | |
| "loss": 3.9072, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.6881695388914766, | |
| "grad_norm": 0.299448162317276, | |
| "learning_rate": 0.000580085032032615, | |
| "loss": 3.8842, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.7027247321844434, | |
| "grad_norm": 0.32242727279663086, | |
| "learning_rate": 0.0005799103086779265, | |
| "loss": 3.8825, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.7172799254774103, | |
| "grad_norm": 0.3023882508277893, | |
| "learning_rate": 0.0005797355853232382, | |
| "loss": 3.8735, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.7318351187703773, | |
| "grad_norm": 0.34403297305107117, | |
| "learning_rate": 0.0005795608619685497, | |
| "loss": 3.8863, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "grad_norm": 0.3128809928894043, | |
| "learning_rate": 0.0005793861386138614, | |
| "loss": 3.8772, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "eval_accuracy": 0.33771024606574407, | |
| "eval_loss": 3.8521087169647217, | |
| "eval_runtime": 179.351, | |
| "eval_samples_per_second": 92.84, | |
| "eval_steps_per_second": 5.804, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.760945505356311, | |
| "grad_norm": 0.32059556245803833, | |
| "learning_rate": 0.0005792114152591729, | |
| "loss": 3.8643, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.775500698649278, | |
| "grad_norm": 0.3130369186401367, | |
| "learning_rate": 0.0005790366919044846, | |
| "loss": 3.8665, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.7900558919422451, | |
| "grad_norm": 0.3213827908039093, | |
| "learning_rate": 0.0005788619685497961, | |
| "loss": 3.8755, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.804611085235212, | |
| "grad_norm": 0.2993370294570923, | |
| "learning_rate": 0.0005786872451951077, | |
| "loss": 3.8674, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.8191662785281788, | |
| "grad_norm": 0.3117568790912628, | |
| "learning_rate": 0.0005785125218404193, | |
| "loss": 3.8717, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.8337214718211459, | |
| "grad_norm": 0.3372628092765808, | |
| "learning_rate": 0.0005783377984857308, | |
| "loss": 3.8741, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.8482766651141127, | |
| "grad_norm": 0.3122817575931549, | |
| "learning_rate": 0.0005781630751310425, | |
| "loss": 3.8657, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.8628318584070795, | |
| "grad_norm": 0.32614654302597046, | |
| "learning_rate": 0.0005779883517763541, | |
| "loss": 3.8523, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.8773870517000466, | |
| "grad_norm": 0.3281755745410919, | |
| "learning_rate": 0.0005778136284216657, | |
| "loss": 3.849, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.8919422449930137, | |
| "grad_norm": 0.32258889079093933, | |
| "learning_rate": 0.0005776389050669772, | |
| "loss": 3.8664, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.9064974382859803, | |
| "grad_norm": 0.3219572603702545, | |
| "learning_rate": 0.0005774641817122889, | |
| "loss": 3.851, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.9210526315789473, | |
| "grad_norm": 0.3125177025794983, | |
| "learning_rate": 0.0005772894583576004, | |
| "loss": 3.8439, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.9356078248719144, | |
| "grad_norm": 0.3299446403980255, | |
| "learning_rate": 0.000577114735002912, | |
| "loss": 3.8437, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.9501630181648812, | |
| "grad_norm": 0.29954883456230164, | |
| "learning_rate": 0.0005769400116482236, | |
| "loss": 3.8372, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.964718211457848, | |
| "grad_norm": 0.3428015112876892, | |
| "learning_rate": 0.0005767652882935352, | |
| "loss": 3.842, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.9792734047508151, | |
| "grad_norm": 0.30913880467414856, | |
| "learning_rate": 0.0005765905649388468, | |
| "loss": 3.8458, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.993828598043782, | |
| "grad_norm": 0.32585445046424866, | |
| "learning_rate": 0.0005764158415841583, | |
| "loss": 3.8499, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 2.0081509082440614, | |
| "grad_norm": 0.33030200004577637, | |
| "learning_rate": 0.00057624111822947, | |
| "loss": 3.7942, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.0227061015370285, | |
| "grad_norm": 0.34908726811408997, | |
| "learning_rate": 0.0005760663948747816, | |
| "loss": 3.7466, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "grad_norm": 0.3279435336589813, | |
| "learning_rate": 0.0005758916715200931, | |
| "loss": 3.7465, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "eval_accuracy": 0.34202185279211383, | |
| "eval_loss": 3.8094260692596436, | |
| "eval_runtime": 179.4152, | |
| "eval_samples_per_second": 92.807, | |
| "eval_steps_per_second": 5.802, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.051816488122962, | |
| "grad_norm": 0.32543864846229553, | |
| "learning_rate": 0.0005757169481654047, | |
| "loss": 3.7445, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 2.066371681415929, | |
| "grad_norm": 0.3198238015174866, | |
| "learning_rate": 0.0005755422248107163, | |
| "loss": 3.7343, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.0809268747088963, | |
| "grad_norm": 0.3203284442424774, | |
| "learning_rate": 0.0005753675014560279, | |
| "loss": 3.7459, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 2.095482068001863, | |
| "grad_norm": 0.3182157576084137, | |
| "learning_rate": 0.0005751927781013395, | |
| "loss": 3.7434, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.11003726129483, | |
| "grad_norm": 0.3111145794391632, | |
| "learning_rate": 0.0005750180547466511, | |
| "loss": 3.7489, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 2.124592454587797, | |
| "grad_norm": 0.3280778229236603, | |
| "learning_rate": 0.0005748433313919627, | |
| "loss": 3.7433, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.139147647880764, | |
| "grad_norm": 0.31418535113334656, | |
| "learning_rate": 0.0005746686080372743, | |
| "loss": 3.7502, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 2.1537028411737307, | |
| "grad_norm": 0.30446282029151917, | |
| "learning_rate": 0.0005744938846825858, | |
| "loss": 3.7584, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.1682580344666977, | |
| "grad_norm": 0.3233334422111511, | |
| "learning_rate": 0.0005743191613278974, | |
| "loss": 3.7633, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 2.182813227759665, | |
| "grad_norm": 0.33291539549827576, | |
| "learning_rate": 0.0005741444379732091, | |
| "loss": 3.7499, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.1973684210526314, | |
| "grad_norm": 0.30665940046310425, | |
| "learning_rate": 0.0005739697146185206, | |
| "loss": 3.7578, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 2.2119236143455985, | |
| "grad_norm": 0.33006513118743896, | |
| "learning_rate": 0.0005737949912638322, | |
| "loss": 3.7439, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.2264788076385655, | |
| "grad_norm": 0.33310437202453613, | |
| "learning_rate": 0.0005736202679091438, | |
| "loss": 3.7458, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 2.241034000931532, | |
| "grad_norm": 0.3161093294620514, | |
| "learning_rate": 0.0005734455445544554, | |
| "loss": 3.7432, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.255589194224499, | |
| "grad_norm": 0.3161999583244324, | |
| "learning_rate": 0.000573270821199767, | |
| "loss": 3.7499, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 2.2701443875174663, | |
| "grad_norm": 0.32372191548347473, | |
| "learning_rate": 0.0005730960978450785, | |
| "loss": 3.7527, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.2846995808104333, | |
| "grad_norm": 0.31958499550819397, | |
| "learning_rate": 0.0005729213744903902, | |
| "loss": 3.7505, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 2.2992547741034, | |
| "grad_norm": 0.3293428421020508, | |
| "learning_rate": 0.0005727466511357017, | |
| "loss": 3.7562, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.313809967396367, | |
| "grad_norm": 0.3318077623844147, | |
| "learning_rate": 0.0005725719277810134, | |
| "loss": 3.7479, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "grad_norm": 0.31161871552467346, | |
| "learning_rate": 0.0005723972044263249, | |
| "loss": 3.7597, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "eval_accuracy": 0.3449051501758032, | |
| "eval_loss": 3.7793445587158203, | |
| "eval_runtime": 179.445, | |
| "eval_samples_per_second": 92.792, | |
| "eval_steps_per_second": 5.801, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.3429203539823007, | |
| "grad_norm": 0.3241608440876007, | |
| "learning_rate": 0.0005722224810716366, | |
| "loss": 3.731, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 2.3574755472752678, | |
| "grad_norm": 0.302915096282959, | |
| "learning_rate": 0.0005720477577169481, | |
| "loss": 3.7498, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.372030740568235, | |
| "grad_norm": 0.30697834491729736, | |
| "learning_rate": 0.0005718730343622598, | |
| "loss": 3.7483, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 2.386585933861202, | |
| "grad_norm": 0.30782854557037354, | |
| "learning_rate": 0.0005716983110075713, | |
| "loss": 3.7506, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.4011411271541685, | |
| "grad_norm": 0.3194926083087921, | |
| "learning_rate": 0.0005715235876528828, | |
| "loss": 3.7471, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 2.4156963204471356, | |
| "grad_norm": 0.32422101497650146, | |
| "learning_rate": 0.0005713488642981945, | |
| "loss": 3.7533, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.4302515137401026, | |
| "grad_norm": 0.3300728499889374, | |
| "learning_rate": 0.0005711741409435061, | |
| "loss": 3.7488, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 2.4448067070330692, | |
| "grad_norm": 0.3104513883590698, | |
| "learning_rate": 0.0005709994175888177, | |
| "loss": 3.7511, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.4593619003260363, | |
| "grad_norm": 0.3395399749279022, | |
| "learning_rate": 0.0005708246942341292, | |
| "loss": 3.755, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 2.4739170936190034, | |
| "grad_norm": 0.321796715259552, | |
| "learning_rate": 0.0005706499708794409, | |
| "loss": 3.7581, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.4884722869119704, | |
| "grad_norm": 0.33986401557922363, | |
| "learning_rate": 0.0005704752475247524, | |
| "loss": 3.7325, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 2.503027480204937, | |
| "grad_norm": 0.3136991262435913, | |
| "learning_rate": 0.0005703005241700641, | |
| "loss": 3.7543, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.517582673497904, | |
| "grad_norm": 0.3213069438934326, | |
| "learning_rate": 0.0005701258008153756, | |
| "loss": 3.7426, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 2.5321378667908707, | |
| "grad_norm": 0.32211968302726746, | |
| "learning_rate": 0.0005699510774606872, | |
| "loss": 3.7248, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 2.546693060083838, | |
| "grad_norm": 0.304557204246521, | |
| "learning_rate": 0.0005697763541059988, | |
| "loss": 3.7378, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 2.561248253376805, | |
| "grad_norm": 0.32637786865234375, | |
| "learning_rate": 0.0005696016307513103, | |
| "loss": 3.7449, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.575803446669772, | |
| "grad_norm": 0.31502121686935425, | |
| "learning_rate": 0.000569426907396622, | |
| "loss": 3.7256, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 2.590358639962739, | |
| "grad_norm": 0.32065874338150024, | |
| "learning_rate": 0.0005692521840419336, | |
| "loss": 3.7397, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 2.6049138332557056, | |
| "grad_norm": 0.3020254671573639, | |
| "learning_rate": 0.0005690774606872452, | |
| "loss": 3.7328, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "grad_norm": 0.32158565521240234, | |
| "learning_rate": 0.0005689027373325567, | |
| "loss": 3.7427, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "eval_accuracy": 0.347388149557257, | |
| "eval_loss": 3.7533369064331055, | |
| "eval_runtime": 179.4903, | |
| "eval_samples_per_second": 92.768, | |
| "eval_steps_per_second": 5.8, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6340242198416393, | |
| "grad_norm": 0.32961833477020264, | |
| "learning_rate": 0.0005687280139778683, | |
| "loss": 3.7421, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 2.6485794131346063, | |
| "grad_norm": 0.34298452734947205, | |
| "learning_rate": 0.0005685532906231799, | |
| "loss": 3.7278, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 2.6631346064275734, | |
| "grad_norm": 0.3249497711658478, | |
| "learning_rate": 0.0005683785672684915, | |
| "loss": 3.7257, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 2.6776897997205404, | |
| "grad_norm": 0.3133349120616913, | |
| "learning_rate": 0.0005682038439138031, | |
| "loss": 3.7441, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.692244993013507, | |
| "grad_norm": 0.300086110830307, | |
| "learning_rate": 0.0005680291205591147, | |
| "loss": 3.7475, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 2.706800186306474, | |
| "grad_norm": 0.3237035870552063, | |
| "learning_rate": 0.0005678543972044263, | |
| "loss": 3.7394, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 2.721355379599441, | |
| "grad_norm": 0.32211583852767944, | |
| "learning_rate": 0.0005676796738497378, | |
| "loss": 3.7323, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 2.735910572892408, | |
| "grad_norm": 0.31221750378608704, | |
| "learning_rate": 0.0005675049504950495, | |
| "loss": 3.7259, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.750465766185375, | |
| "grad_norm": 0.3241555690765381, | |
| "learning_rate": 0.0005673302271403611, | |
| "loss": 3.7383, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 2.765020959478342, | |
| "grad_norm": 0.29803764820098877, | |
| "learning_rate": 0.0005671555037856726, | |
| "loss": 3.7216, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.779576152771309, | |
| "grad_norm": 0.3058941662311554, | |
| "learning_rate": 0.0005669807804309842, | |
| "loss": 3.7273, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 2.7941313460642756, | |
| "grad_norm": 0.3076792061328888, | |
| "learning_rate": 0.0005668060570762958, | |
| "loss": 3.729, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 2.8086865393572427, | |
| "grad_norm": 0.3100437819957733, | |
| "learning_rate": 0.0005666313337216074, | |
| "loss": 3.7223, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 2.8232417326502097, | |
| "grad_norm": 0.312364786863327, | |
| "learning_rate": 0.000566456610366919, | |
| "loss": 3.7305, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 2.8377969259431763, | |
| "grad_norm": 0.31409692764282227, | |
| "learning_rate": 0.0005662818870122306, | |
| "loss": 3.7071, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 2.8523521192361434, | |
| "grad_norm": 0.30859071016311646, | |
| "learning_rate": 0.0005661071636575422, | |
| "loss": 3.7384, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 2.8669073125291105, | |
| "grad_norm": 0.3208765685558319, | |
| "learning_rate": 0.0005659324403028537, | |
| "loss": 3.7277, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 2.8814625058220775, | |
| "grad_norm": 0.29655808210372925, | |
| "learning_rate": 0.0005657577169481653, | |
| "loss": 3.7154, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 2.896017699115044, | |
| "grad_norm": 0.3313380777835846, | |
| "learning_rate": 0.0005655829935934769, | |
| "loss": 3.7181, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "grad_norm": 0.2961195707321167, | |
| "learning_rate": 0.0005654082702387886, | |
| "loss": 3.7251, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "eval_accuracy": 0.34988078617775464, | |
| "eval_loss": 3.723102331161499, | |
| "eval_runtime": 179.4175, | |
| "eval_samples_per_second": 92.806, | |
| "eval_steps_per_second": 5.802, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.9251280857009783, | |
| "grad_norm": 0.30046647787094116, | |
| "learning_rate": 0.0005652335468841001, | |
| "loss": 3.6131, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 2.939683278993945, | |
| "grad_norm": 0.33831682801246643, | |
| "learning_rate": 0.0005650588235294117, | |
| "loss": 3.6069, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 2.954238472286912, | |
| "grad_norm": 0.3195353150367737, | |
| "learning_rate": 0.0005648841001747233, | |
| "loss": 3.6146, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 2.968793665579879, | |
| "grad_norm": 0.30412888526916504, | |
| "learning_rate": 0.0005647093768200349, | |
| "loss": 3.6043, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 2.983348858872846, | |
| "grad_norm": 0.3138900399208069, | |
| "learning_rate": 0.0005645346534653465, | |
| "loss": 3.6143, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 2.9979040521658127, | |
| "grad_norm": 0.31442713737487793, | |
| "learning_rate": 0.0005643599301106582, | |
| "loss": 3.6146, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 3.0125174662319516, | |
| "grad_norm": 0.33579516410827637, | |
| "learning_rate": 0.0005641852067559697, | |
| "loss": 3.6742, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 3.0270726595249187, | |
| "grad_norm": 0.32205700874328613, | |
| "learning_rate": 0.0005640104834012812, | |
| "loss": 3.6072, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 3.0416278528178853, | |
| "grad_norm": 0.3189714550971985, | |
| "learning_rate": 0.0005638357600465929, | |
| "loss": 3.6164, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 3.0561830461108523, | |
| "grad_norm": 0.3315085172653198, | |
| "learning_rate": 0.0005636610366919044, | |
| "loss": 3.6037, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.0707382394038194, | |
| "grad_norm": 0.3438396155834198, | |
| "learning_rate": 0.0005634863133372161, | |
| "loss": 3.6141, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 3.085293432696786, | |
| "grad_norm": 0.32832789421081543, | |
| "learning_rate": 0.0005633115899825276, | |
| "loss": 3.6291, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 3.099848625989753, | |
| "grad_norm": 0.33390843868255615, | |
| "learning_rate": 0.0005631368666278393, | |
| "loss": 3.6353, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 3.11440381928272, | |
| "grad_norm": 0.32860368490219116, | |
| "learning_rate": 0.0005629621432731508, | |
| "loss": 3.628, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 3.128959012575687, | |
| "grad_norm": 0.32481637597084045, | |
| "learning_rate": 0.0005627874199184623, | |
| "loss": 3.6332, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 3.143514205868654, | |
| "grad_norm": 0.32326942682266235, | |
| "learning_rate": 0.000562612696563774, | |
| "loss": 3.6268, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 3.158069399161621, | |
| "grad_norm": 0.3113882541656494, | |
| "learning_rate": 0.0005624379732090856, | |
| "loss": 3.6337, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 3.172624592454588, | |
| "grad_norm": 0.32441622018814087, | |
| "learning_rate": 0.0005622632498543972, | |
| "loss": 3.6364, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 3.1871797857475546, | |
| "grad_norm": 0.31897199153900146, | |
| "learning_rate": 0.0005620885264997087, | |
| "loss": 3.6257, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 3.2017349790405216, | |
| "grad_norm": 0.3312537968158722, | |
| "learning_rate": 0.0005619138031450204, | |
| "loss": 3.629, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2017349790405216, | |
| "eval_accuracy": 0.3520870087549616, | |
| "eval_loss": 3.7162582874298096, | |
| "eval_runtime": 181.7923, | |
| "eval_samples_per_second": 91.594, | |
| "eval_steps_per_second": 5.726, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2162901723334887, | |
| "grad_norm": 0.32515862584114075, | |
| "learning_rate": 0.0005617390797903319, | |
| "loss": 3.6315, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 3.2308453656264557, | |
| "grad_norm": 0.2998684048652649, | |
| "learning_rate": 0.0005615643564356436, | |
| "loss": 3.64, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 3.2454005589194224, | |
| "grad_norm": 0.3279356062412262, | |
| "learning_rate": 0.0005613896330809551, | |
| "loss": 3.6487, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 3.2599557522123894, | |
| "grad_norm": 0.31287768483161926, | |
| "learning_rate": 0.0005612149097262667, | |
| "loss": 3.6286, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 3.2745109455053565, | |
| "grad_norm": 0.34031912684440613, | |
| "learning_rate": 0.0005610401863715783, | |
| "loss": 3.6535, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 3.289066138798323, | |
| "grad_norm": 0.32775530219078064, | |
| "learning_rate": 0.0005608654630168898, | |
| "loss": 3.6403, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 3.30362133209129, | |
| "grad_norm": 0.30365148186683655, | |
| "learning_rate": 0.0005606907396622015, | |
| "loss": 3.6398, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 3.3181765253842572, | |
| "grad_norm": 0.31511229276657104, | |
| "learning_rate": 0.0005605160163075131, | |
| "loss": 3.6436, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 3.332731718677224, | |
| "grad_norm": 0.3300502896308899, | |
| "learning_rate": 0.0005603412929528247, | |
| "loss": 3.6464, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 3.347286911970191, | |
| "grad_norm": 0.3103245198726654, | |
| "learning_rate": 0.0005601665695981362, | |
| "loss": 3.636, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.361842105263158, | |
| "grad_norm": 0.322672039270401, | |
| "learning_rate": 0.0005599918462434478, | |
| "loss": 3.6354, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 3.3763972985561246, | |
| "grad_norm": 0.3106880486011505, | |
| "learning_rate": 0.0005598171228887594, | |
| "loss": 3.6439, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 3.3909524918490916, | |
| "grad_norm": 0.3212052583694458, | |
| "learning_rate": 0.0005596423995340709, | |
| "loss": 3.6357, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 3.4055076851420587, | |
| "grad_norm": 0.3229770362377167, | |
| "learning_rate": 0.0005594676761793826, | |
| "loss": 3.6416, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 3.4200628784350258, | |
| "grad_norm": 0.31980758905410767, | |
| "learning_rate": 0.0005592929528246942, | |
| "loss": 3.6531, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 3.4346180717279924, | |
| "grad_norm": 0.3255523145198822, | |
| "learning_rate": 0.0005591182294700058, | |
| "loss": 3.6435, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 3.4491732650209594, | |
| "grad_norm": 0.3042266368865967, | |
| "learning_rate": 0.0005589435061153173, | |
| "loss": 3.6418, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 3.4637284583139265, | |
| "grad_norm": 0.32445093989372253, | |
| "learning_rate": 0.000558768782760629, | |
| "loss": 3.6515, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 3.478283651606893, | |
| "grad_norm": 0.31971192359924316, | |
| "learning_rate": 0.0005585940594059406, | |
| "loss": 3.6484, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 3.49283884489986, | |
| "grad_norm": 0.3065326511859894, | |
| "learning_rate": 0.0005584193360512521, | |
| "loss": 3.6484, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.49283884489986, | |
| "eval_accuracy": 0.35326992108394023, | |
| "eval_loss": 3.6962692737579346, | |
| "eval_runtime": 180.8346, | |
| "eval_samples_per_second": 92.079, | |
| "eval_steps_per_second": 5.757, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.5073940381928272, | |
| "grad_norm": 0.3139113485813141, | |
| "learning_rate": 0.0005582446126965637, | |
| "loss": 3.6446, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 3.5219492314857943, | |
| "grad_norm": 0.32031387090682983, | |
| "learning_rate": 0.0005580698893418753, | |
| "loss": 3.6495, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 3.536504424778761, | |
| "grad_norm": 0.3141990602016449, | |
| "learning_rate": 0.0005578951659871869, | |
| "loss": 3.6398, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 3.551059618071728, | |
| "grad_norm": 0.3226875066757202, | |
| "learning_rate": 0.0005577204426324985, | |
| "loss": 3.6488, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 3.565614811364695, | |
| "grad_norm": 0.30300959944725037, | |
| "learning_rate": 0.0005575457192778101, | |
| "loss": 3.6313, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 3.5801700046576617, | |
| "grad_norm": 0.3211573660373688, | |
| "learning_rate": 0.0005573709959231217, | |
| "loss": 3.6463, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 3.5947251979506287, | |
| "grad_norm": 0.31371423602104187, | |
| "learning_rate": 0.0005571962725684332, | |
| "loss": 3.6475, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 3.609280391243596, | |
| "grad_norm": 0.31352680921554565, | |
| "learning_rate": 0.0005570215492137449, | |
| "loss": 3.6435, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 3.623835584536563, | |
| "grad_norm": 0.29655176401138306, | |
| "learning_rate": 0.0005568468258590564, | |
| "loss": 3.6431, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 3.6383907778295295, | |
| "grad_norm": 0.32901984453201294, | |
| "learning_rate": 0.0005566721025043681, | |
| "loss": 3.632, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.6529459711224965, | |
| "grad_norm": 0.29682856798171997, | |
| "learning_rate": 0.0005564973791496796, | |
| "loss": 3.649, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 3.6675011644154636, | |
| "grad_norm": 0.32371819019317627, | |
| "learning_rate": 0.0005563226557949913, | |
| "loss": 3.6322, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 3.68205635770843, | |
| "grad_norm": 0.3183552324771881, | |
| "learning_rate": 0.0005561479324403028, | |
| "loss": 3.6362, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 3.6966115510013973, | |
| "grad_norm": 0.31535252928733826, | |
| "learning_rate": 0.0005559732090856144, | |
| "loss": 3.6283, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 3.7111667442943643, | |
| "grad_norm": 0.30772748589515686, | |
| "learning_rate": 0.000555798485730926, | |
| "loss": 3.6552, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 3.7257219375873314, | |
| "grad_norm": 0.32342010736465454, | |
| "learning_rate": 0.0005556237623762376, | |
| "loss": 3.6481, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 3.740277130880298, | |
| "grad_norm": 0.3001164495944977, | |
| "learning_rate": 0.0005554490390215492, | |
| "loss": 3.632, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 3.754832324173265, | |
| "grad_norm": 0.3260549306869507, | |
| "learning_rate": 0.0005552743156668607, | |
| "loss": 3.6338, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 3.7693875174662317, | |
| "grad_norm": 0.325056254863739, | |
| "learning_rate": 0.0005550995923121724, | |
| "loss": 3.6336, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 3.7839427107591987, | |
| "grad_norm": 0.3091551959514618, | |
| "learning_rate": 0.0005549248689574839, | |
| "loss": 3.6332, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.7839427107591987, | |
| "eval_accuracy": 0.354759579680046, | |
| "eval_loss": 3.6787266731262207, | |
| "eval_runtime": 180.7938, | |
| "eval_samples_per_second": 92.099, | |
| "eval_steps_per_second": 5.758, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.798497904052166, | |
| "grad_norm": 0.3215680718421936, | |
| "learning_rate": 0.0005547501456027955, | |
| "loss": 3.646, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 3.813053097345133, | |
| "grad_norm": 0.30009904503822327, | |
| "learning_rate": 0.0005545754222481071, | |
| "loss": 3.6497, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 3.8276082906381, | |
| "grad_norm": 0.31329527497291565, | |
| "learning_rate": 0.0005544006988934188, | |
| "loss": 3.6408, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 3.8421634839310665, | |
| "grad_norm": 0.3503079116344452, | |
| "learning_rate": 0.0005542259755387303, | |
| "loss": 3.6464, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 3.8567186772240336, | |
| "grad_norm": 0.32458430528640747, | |
| "learning_rate": 0.0005540512521840418, | |
| "loss": 3.6416, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 3.8712738705170002, | |
| "grad_norm": 0.28616443276405334, | |
| "learning_rate": 0.0005538765288293535, | |
| "loss": 3.6359, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 3.8858290638099673, | |
| "grad_norm": 0.310008704662323, | |
| "learning_rate": 0.0005537018054746651, | |
| "loss": 3.6406, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 3.9003842571029343, | |
| "grad_norm": 0.3077920377254486, | |
| "learning_rate": 0.0005535270821199767, | |
| "loss": 3.6404, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 3.9149394503959014, | |
| "grad_norm": 0.3122752904891968, | |
| "learning_rate": 0.0005533523587652882, | |
| "loss": 3.6297, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 3.929494643688868, | |
| "grad_norm": 0.30146297812461853, | |
| "learning_rate": 0.0005531776354105999, | |
| "loss": 3.6378, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.944049836981835, | |
| "grad_norm": 0.3015786409378052, | |
| "learning_rate": 0.0005530029120559114, | |
| "loss": 3.6414, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 3.958605030274802, | |
| "grad_norm": 0.3123229444026947, | |
| "learning_rate": 0.0005528281887012229, | |
| "loss": 3.64, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 3.9731602235677688, | |
| "grad_norm": 0.3401547968387604, | |
| "learning_rate": 0.0005526534653465346, | |
| "loss": 3.6367, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 3.987715416860736, | |
| "grad_norm": 0.327626496553421, | |
| "learning_rate": 0.0005524787419918462, | |
| "loss": 3.6363, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 4.0020377270610155, | |
| "grad_norm": 0.3363268971443176, | |
| "learning_rate": 0.0005523040186371578, | |
| "loss": 3.6228, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 4.0165929203539825, | |
| "grad_norm": 0.3198484480381012, | |
| "learning_rate": 0.0005521292952824693, | |
| "loss": 3.5255, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 4.03114811364695, | |
| "grad_norm": 0.3151451349258423, | |
| "learning_rate": 0.000551954571927781, | |
| "loss": 3.5266, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 4.045703306939916, | |
| "grad_norm": 0.3302767872810364, | |
| "learning_rate": 0.0005517798485730926, | |
| "loss": 3.5197, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 4.060258500232883, | |
| "grad_norm": 0.3168678283691406, | |
| "learning_rate": 0.0005516051252184042, | |
| "loss": 3.528, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 4.07481369352585, | |
| "grad_norm": 0.33349886536598206, | |
| "learning_rate": 0.0005514304018637157, | |
| "loss": 3.5315, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.07481369352585, | |
| "eval_accuracy": 0.3561975262617702, | |
| "eval_loss": 3.6728265285491943, | |
| "eval_runtime": 180.7141, | |
| "eval_samples_per_second": 92.14, | |
| "eval_steps_per_second": 5.76, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.089368886818817, | |
| "grad_norm": 0.32183805108070374, | |
| "learning_rate": 0.0005512556785090273, | |
| "loss": 3.5357, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 4.103924080111784, | |
| "grad_norm": 0.31829333305358887, | |
| "learning_rate": 0.0005510809551543389, | |
| "loss": 3.5455, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 4.118479273404751, | |
| "grad_norm": 0.30199795961380005, | |
| "learning_rate": 0.0005509062317996504, | |
| "loss": 3.5482, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 4.133034466697718, | |
| "grad_norm": 0.311040461063385, | |
| "learning_rate": 0.0005507315084449621, | |
| "loss": 3.5449, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 4.147589659990684, | |
| "grad_norm": 0.33546823263168335, | |
| "learning_rate": 0.0005505567850902737, | |
| "loss": 3.5483, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 4.162144853283651, | |
| "grad_norm": 0.33755001425743103, | |
| "learning_rate": 0.0005503820617355853, | |
| "loss": 3.5585, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 4.176700046576618, | |
| "grad_norm": 0.3277694582939148, | |
| "learning_rate": 0.0005502073383808969, | |
| "loss": 3.566, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 4.1912552398695855, | |
| "grad_norm": 0.32476261258125305, | |
| "learning_rate": 0.0005500326150262085, | |
| "loss": 3.551, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 4.2058104331625525, | |
| "grad_norm": 0.35436880588531494, | |
| "learning_rate": 0.00054985789167152, | |
| "loss": 3.5592, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 4.22036562645552, | |
| "grad_norm": 0.32852813601493835, | |
| "learning_rate": 0.0005496831683168316, | |
| "loss": 3.5695, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.234920819748487, | |
| "grad_norm": 0.31876441836357117, | |
| "learning_rate": 0.0005495084449621433, | |
| "loss": 3.5533, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 4.249476013041453, | |
| "grad_norm": 0.3397602140903473, | |
| "learning_rate": 0.0005493337216074548, | |
| "loss": 3.5484, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 4.26403120633442, | |
| "grad_norm": 0.30222707986831665, | |
| "learning_rate": 0.0005491589982527664, | |
| "loss": 3.5632, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 4.278586399627387, | |
| "grad_norm": 0.3110881745815277, | |
| "learning_rate": 0.000548984274898078, | |
| "loss": 3.5593, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 4.293141592920354, | |
| "grad_norm": 0.32440635561943054, | |
| "learning_rate": 0.0005488095515433897, | |
| "loss": 3.556, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 4.307696786213321, | |
| "grad_norm": 0.31460386514663696, | |
| "learning_rate": 0.0005486348281887012, | |
| "loss": 3.567, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 4.322251979506288, | |
| "grad_norm": 0.2991078495979309, | |
| "learning_rate": 0.0005484601048340127, | |
| "loss": 3.563, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 4.336807172799254, | |
| "grad_norm": 0.343728631734848, | |
| "learning_rate": 0.0005482853814793244, | |
| "loss": 3.5608, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 4.351362366092221, | |
| "grad_norm": 0.32072120904922485, | |
| "learning_rate": 0.0005481106581246359, | |
| "loss": 3.5685, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 4.365917559385188, | |
| "grad_norm": 0.3259231448173523, | |
| "learning_rate": 0.0005479359347699475, | |
| "loss": 3.5681, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.365917559385188, | |
| "eval_accuracy": 0.3573054561698956, | |
| "eval_loss": 3.6608378887176514, | |
| "eval_runtime": 180.9212, | |
| "eval_samples_per_second": 92.035, | |
| "eval_steps_per_second": 5.754, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.3804727526781555, | |
| "grad_norm": 0.3365843892097473, | |
| "learning_rate": 0.0005477612114152591, | |
| "loss": 3.5708, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 4.395027945971123, | |
| "grad_norm": 0.32022547721862793, | |
| "learning_rate": 0.0005475864880605708, | |
| "loss": 3.5736, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 4.40958313926409, | |
| "grad_norm": 0.3090197741985321, | |
| "learning_rate": 0.0005474117647058823, | |
| "loss": 3.5685, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 4.424138332557057, | |
| "grad_norm": 0.31479737162590027, | |
| "learning_rate": 0.0005472370413511939, | |
| "loss": 3.5693, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 4.438693525850024, | |
| "grad_norm": 0.3122880458831787, | |
| "learning_rate": 0.0005470623179965055, | |
| "loss": 3.5662, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 4.45324871914299, | |
| "grad_norm": 0.31671661138534546, | |
| "learning_rate": 0.0005468875946418171, | |
| "loss": 3.5729, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 4.467803912435957, | |
| "grad_norm": 0.32296523451805115, | |
| "learning_rate": 0.0005467128712871287, | |
| "loss": 3.5622, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 4.482359105728924, | |
| "grad_norm": 0.315123975276947, | |
| "learning_rate": 0.0005465381479324402, | |
| "loss": 3.5623, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 4.496914299021891, | |
| "grad_norm": 0.3224656581878662, | |
| "learning_rate": 0.0005463634245777519, | |
| "loss": 3.5768, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 4.511469492314858, | |
| "grad_norm": 0.329997181892395, | |
| "learning_rate": 0.0005461887012230634, | |
| "loss": 3.5794, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 4.526024685607825, | |
| "grad_norm": 0.3304203748703003, | |
| "learning_rate": 0.000546013977868375, | |
| "loss": 3.5778, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 4.540579878900791, | |
| "grad_norm": 0.3303874433040619, | |
| "learning_rate": 0.0005458392545136866, | |
| "loss": 3.5719, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 4.5551350721937585, | |
| "grad_norm": 0.31413403153419495, | |
| "learning_rate": 0.0005456645311589983, | |
| "loss": 3.5651, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 4.5696902654867255, | |
| "grad_norm": 0.3068625032901764, | |
| "learning_rate": 0.0005454898078043098, | |
| "loss": 3.5817, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 4.584245458779693, | |
| "grad_norm": 0.30968910455703735, | |
| "learning_rate": 0.0005453150844496213, | |
| "loss": 3.5769, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 4.59880065207266, | |
| "grad_norm": 0.34386196732521057, | |
| "learning_rate": 0.000545140361094933, | |
| "loss": 3.581, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 4.613355845365627, | |
| "grad_norm": 0.32992249727249146, | |
| "learning_rate": 0.0005449656377402445, | |
| "loss": 3.5871, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 4.627911038658594, | |
| "grad_norm": 0.31546783447265625, | |
| "learning_rate": 0.0005447909143855562, | |
| "loss": 3.5648, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 4.64246623195156, | |
| "grad_norm": 0.3235871493816376, | |
| "learning_rate": 0.0005446161910308677, | |
| "loss": 3.5675, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 4.657021425244527, | |
| "grad_norm": 0.32952284812927246, | |
| "learning_rate": 0.0005444414676761794, | |
| "loss": 3.5781, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.657021425244527, | |
| "eval_accuracy": 0.3587489265349742, | |
| "eval_loss": 3.6463191509246826, | |
| "eval_runtime": 180.9022, | |
| "eval_samples_per_second": 92.044, | |
| "eval_steps_per_second": 5.754, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.671576618537494, | |
| "grad_norm": 0.2940504848957062, | |
| "learning_rate": 0.0005442667443214909, | |
| "loss": 3.583, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 4.686131811830461, | |
| "grad_norm": 0.3155236840248108, | |
| "learning_rate": 0.0005440920209668024, | |
| "loss": 3.5748, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 4.700687005123428, | |
| "grad_norm": 0.30486825108528137, | |
| "learning_rate": 0.0005439172976121141, | |
| "loss": 3.5818, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 4.715242198416395, | |
| "grad_norm": 0.3136492669582367, | |
| "learning_rate": 0.0005437425742574257, | |
| "loss": 3.5634, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 4.729797391709361, | |
| "grad_norm": 0.30183354020118713, | |
| "learning_rate": 0.0005435678509027373, | |
| "loss": 3.5841, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 4.7443525850023285, | |
| "grad_norm": 0.3134796619415283, | |
| "learning_rate": 0.0005433931275480488, | |
| "loss": 3.5798, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 4.7589077782952955, | |
| "grad_norm": 0.3129316568374634, | |
| "learning_rate": 0.0005432184041933605, | |
| "loss": 3.5788, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 4.773462971588263, | |
| "grad_norm": 0.3630964457988739, | |
| "learning_rate": 0.000543043680838672, | |
| "loss": 3.5812, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 4.78801816488123, | |
| "grad_norm": 0.3073498606681824, | |
| "learning_rate": 0.0005428689574839837, | |
| "loss": 3.5714, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 4.802573358174197, | |
| "grad_norm": 0.31628504395484924, | |
| "learning_rate": 0.0005426942341292952, | |
| "loss": 3.5832, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 4.817128551467164, | |
| "grad_norm": 0.3307545483112335, | |
| "learning_rate": 0.0005425195107746068, | |
| "loss": 3.577, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 4.831683744760131, | |
| "grad_norm": 0.30894699692726135, | |
| "learning_rate": 0.0005423447874199184, | |
| "loss": 3.5857, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 4.846238938053097, | |
| "grad_norm": 0.3408315181732178, | |
| "learning_rate": 0.00054217006406523, | |
| "loss": 3.57, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 4.860794131346064, | |
| "grad_norm": 0.3083992600440979, | |
| "learning_rate": 0.0005419953407105417, | |
| "loss": 3.5818, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 4.875349324639031, | |
| "grad_norm": 0.29149574041366577, | |
| "learning_rate": 0.0005418206173558532, | |
| "loss": 3.5844, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 4.889904517931998, | |
| "grad_norm": 0.31696999073028564, | |
| "learning_rate": 0.0005416458940011648, | |
| "loss": 3.5859, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 4.904459711224965, | |
| "grad_norm": 0.30741801857948303, | |
| "learning_rate": 0.0005414711706464764, | |
| "loss": 3.5762, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 4.919014904517932, | |
| "grad_norm": 0.3022885322570801, | |
| "learning_rate": 0.000541296447291788, | |
| "loss": 3.5817, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 4.9335700978108985, | |
| "grad_norm": 0.3171398937702179, | |
| "learning_rate": 0.0005411217239370995, | |
| "loss": 3.5774, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 4.948125291103866, | |
| "grad_norm": 0.29663971066474915, | |
| "learning_rate": 0.0005409470005824111, | |
| "loss": 3.5613, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.948125291103866, | |
| "eval_accuracy": 0.3598574440796266, | |
| "eval_loss": 3.633816957473755, | |
| "eval_runtime": 180.8783, | |
| "eval_samples_per_second": 92.056, | |
| "eval_steps_per_second": 5.755, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.962680484396833, | |
| "grad_norm": 0.3096961975097656, | |
| "learning_rate": 0.0005407722772277228, | |
| "loss": 3.5658, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 4.9772356776898, | |
| "grad_norm": 0.31567344069480896, | |
| "learning_rate": 0.0005405975538730343, | |
| "loss": 3.5664, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 4.991790870982767, | |
| "grad_norm": 0.35638725757598877, | |
| "learning_rate": 0.0005404228305183459, | |
| "loss": 3.5741, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 5.006113181183046, | |
| "grad_norm": 0.3229467272758484, | |
| "learning_rate": 0.0005402481071636575, | |
| "loss": 3.5306, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 5.020668374476013, | |
| "grad_norm": 0.32992422580718994, | |
| "learning_rate": 0.0005400733838089692, | |
| "loss": 3.4672, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 5.03522356776898, | |
| "grad_norm": 0.3403879404067993, | |
| "learning_rate": 0.0005398986604542807, | |
| "loss": 3.4667, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 5.049778761061947, | |
| "grad_norm": 0.32500892877578735, | |
| "learning_rate": 0.0005397239370995922, | |
| "loss": 3.4752, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 5.064333954354914, | |
| "grad_norm": 0.2997499108314514, | |
| "learning_rate": 0.0005395492137449039, | |
| "loss": 3.4724, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 5.078889147647881, | |
| "grad_norm": 0.3481856882572174, | |
| "learning_rate": 0.0005393744903902154, | |
| "loss": 3.4825, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 5.093444340940848, | |
| "grad_norm": 0.338044136762619, | |
| "learning_rate": 0.000539199767035527, | |
| "loss": 3.4807, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 5.107999534233815, | |
| "grad_norm": 0.3250182867050171, | |
| "learning_rate": 0.0005390250436808386, | |
| "loss": 3.4797, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 5.122554727526782, | |
| "grad_norm": 0.3101995885372162, | |
| "learning_rate": 0.0005388503203261503, | |
| "loss": 3.5014, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 5.137109920819748, | |
| "grad_norm": 0.32210931181907654, | |
| "learning_rate": 0.0005386755969714618, | |
| "loss": 3.4837, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 5.151665114112715, | |
| "grad_norm": 0.32669687271118164, | |
| "learning_rate": 0.0005385008736167733, | |
| "loss": 3.4861, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 5.166220307405682, | |
| "grad_norm": 0.333234041929245, | |
| "learning_rate": 0.000538326150262085, | |
| "loss": 3.4951, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 5.180775500698649, | |
| "grad_norm": 0.3409346640110016, | |
| "learning_rate": 0.0005381514269073965, | |
| "loss": 3.4741, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 5.195330693991616, | |
| "grad_norm": 0.3232422471046448, | |
| "learning_rate": 0.0005379767035527082, | |
| "loss": 3.4955, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 5.2098858872845835, | |
| "grad_norm": 0.32606247067451477, | |
| "learning_rate": 0.0005378019801980197, | |
| "loss": 3.5095, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 5.2244410805775505, | |
| "grad_norm": 0.3317588269710541, | |
| "learning_rate": 0.0005376272568433314, | |
| "loss": 3.4932, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 5.238996273870517, | |
| "grad_norm": 0.3223241865634918, | |
| "learning_rate": 0.0005374525334886429, | |
| "loss": 3.4987, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.238996273870517, | |
| "eval_accuracy": 0.3603531742538573, | |
| "eval_loss": 3.6327548027038574, | |
| "eval_runtime": 181.0812, | |
| "eval_samples_per_second": 91.953, | |
| "eval_steps_per_second": 5.749, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.253551467163484, | |
| "grad_norm": 0.3261396884918213, | |
| "learning_rate": 0.0005372778101339545, | |
| "loss": 3.5104, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 5.268106660456451, | |
| "grad_norm": 0.31246986985206604, | |
| "learning_rate": 0.0005371030867792661, | |
| "loss": 3.5021, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 5.282661853749418, | |
| "grad_norm": 0.320519357919693, | |
| "learning_rate": 0.0005369283634245778, | |
| "loss": 3.4982, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 5.297217047042385, | |
| "grad_norm": 0.3356594443321228, | |
| "learning_rate": 0.0005367536400698893, | |
| "loss": 3.5051, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 5.311772240335352, | |
| "grad_norm": 0.32634884119033813, | |
| "learning_rate": 0.0005365789167152008, | |
| "loss": 3.517, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 5.326327433628318, | |
| "grad_norm": 0.31914022564888, | |
| "learning_rate": 0.0005364041933605125, | |
| "loss": 3.5203, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 5.340882626921285, | |
| "grad_norm": 0.3101184070110321, | |
| "learning_rate": 0.000536229470005824, | |
| "loss": 3.5101, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 5.355437820214252, | |
| "grad_norm": 0.313850075006485, | |
| "learning_rate": 0.0005360547466511357, | |
| "loss": 3.5077, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 5.369993013507219, | |
| "grad_norm": 0.2964019477367401, | |
| "learning_rate": 0.0005358800232964472, | |
| "loss": 3.5284, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 5.384548206800186, | |
| "grad_norm": 0.320024311542511, | |
| "learning_rate": 0.0005357052999417589, | |
| "loss": 3.5195, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 5.3991034000931535, | |
| "grad_norm": 0.3183858394622803, | |
| "learning_rate": 0.0005355305765870704, | |
| "loss": 3.503, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 5.4136585933861205, | |
| "grad_norm": 0.3233298659324646, | |
| "learning_rate": 0.000535355853232382, | |
| "loss": 3.5258, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 5.428213786679087, | |
| "grad_norm": 0.3176547586917877, | |
| "learning_rate": 0.0005351811298776936, | |
| "loss": 3.5182, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 5.442768979972054, | |
| "grad_norm": 0.32504311203956604, | |
| "learning_rate": 0.0005350064065230052, | |
| "loss": 3.5212, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 5.457324173265021, | |
| "grad_norm": 0.32569703459739685, | |
| "learning_rate": 0.0005348316831683168, | |
| "loss": 3.5192, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 5.471879366557988, | |
| "grad_norm": 0.31813210248947144, | |
| "learning_rate": 0.0005346569598136284, | |
| "loss": 3.516, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 5.486434559850955, | |
| "grad_norm": 0.32232987880706787, | |
| "learning_rate": 0.00053448223645894, | |
| "loss": 3.5212, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 5.500989753143922, | |
| "grad_norm": 0.33343178033828735, | |
| "learning_rate": 0.0005343075131042515, | |
| "loss": 3.5267, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 5.515544946436888, | |
| "grad_norm": 0.33466091752052307, | |
| "learning_rate": 0.0005341327897495632, | |
| "loss": 3.5215, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 5.530100139729855, | |
| "grad_norm": 0.3394691050052643, | |
| "learning_rate": 0.0005339580663948748, | |
| "loss": 3.5154, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.530100139729855, | |
| "eval_accuracy": 0.36140610138304957, | |
| "eval_loss": 3.622549533843994, | |
| "eval_runtime": 180.8942, | |
| "eval_samples_per_second": 92.048, | |
| "eval_steps_per_second": 5.755, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.544655333022822, | |
| "grad_norm": 0.31924381852149963, | |
| "learning_rate": 0.0005337833430401863, | |
| "loss": 3.5189, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 5.559210526315789, | |
| "grad_norm": 0.30619359016418457, | |
| "learning_rate": 0.0005336086196854979, | |
| "loss": 3.522, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 5.573765719608756, | |
| "grad_norm": 0.3263663947582245, | |
| "learning_rate": 0.0005334338963308095, | |
| "loss": 3.5163, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 5.5883209129017235, | |
| "grad_norm": 0.32492828369140625, | |
| "learning_rate": 0.0005332591729761211, | |
| "loss": 3.5419, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 5.602876106194691, | |
| "grad_norm": 0.30963876843452454, | |
| "learning_rate": 0.0005330844496214327, | |
| "loss": 3.5158, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 5.617431299487658, | |
| "grad_norm": 0.33830997347831726, | |
| "learning_rate": 0.0005329097262667443, | |
| "loss": 3.5281, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 5.631986492780624, | |
| "grad_norm": 0.3114701211452484, | |
| "learning_rate": 0.0005327350029120559, | |
| "loss": 3.5288, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 5.646541686073591, | |
| "grad_norm": 0.32258859276771545, | |
| "learning_rate": 0.0005325602795573674, | |
| "loss": 3.5321, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 5.661096879366558, | |
| "grad_norm": 0.342039555311203, | |
| "learning_rate": 0.000532385556202679, | |
| "loss": 3.5308, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 5.675652072659525, | |
| "grad_norm": 0.3170706629753113, | |
| "learning_rate": 0.0005322108328479906, | |
| "loss": 3.5177, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 5.690207265952492, | |
| "grad_norm": 0.3374808728694916, | |
| "learning_rate": 0.0005320361094933023, | |
| "loss": 3.5252, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 5.704762459245459, | |
| "grad_norm": 0.29898542165756226, | |
| "learning_rate": 0.0005318613861386138, | |
| "loss": 3.5209, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 5.719317652538425, | |
| "grad_norm": 0.3294072151184082, | |
| "learning_rate": 0.0005316866627839254, | |
| "loss": 3.5228, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 5.733872845831392, | |
| "grad_norm": 0.30430126190185547, | |
| "learning_rate": 0.000531511939429237, | |
| "loss": 3.5207, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 5.748428039124359, | |
| "grad_norm": 0.33038702607154846, | |
| "learning_rate": 0.0005313372160745486, | |
| "loss": 3.5294, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 5.7629832324173265, | |
| "grad_norm": 0.3005201518535614, | |
| "learning_rate": 0.0005311624927198602, | |
| "loss": 3.5347, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 5.7775384257102935, | |
| "grad_norm": 0.3165362775325775, | |
| "learning_rate": 0.0005309877693651717, | |
| "loss": 3.5212, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 5.792093619003261, | |
| "grad_norm": 0.3144179880619049, | |
| "learning_rate": 0.0005308130460104834, | |
| "loss": 3.5243, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 5.806648812296228, | |
| "grad_norm": 0.3042346239089966, | |
| "learning_rate": 0.0005306383226557949, | |
| "loss": 3.5196, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 5.821204005589195, | |
| "grad_norm": 0.3238431513309479, | |
| "learning_rate": 0.0005304635993011065, | |
| "loss": 3.528, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.821204005589195, | |
| "eval_accuracy": 0.36223972256034176, | |
| "eval_loss": 3.6133012771606445, | |
| "eval_runtime": 180.9991, | |
| "eval_samples_per_second": 91.995, | |
| "eval_steps_per_second": 5.751, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.835759198882161, | |
| "grad_norm": 0.32478225231170654, | |
| "learning_rate": 0.0005302888759464181, | |
| "loss": 3.5305, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 5.850314392175128, | |
| "grad_norm": 0.3156353831291199, | |
| "learning_rate": 0.0005301141525917298, | |
| "loss": 3.524, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 5.864869585468095, | |
| "grad_norm": 0.3088098466396332, | |
| "learning_rate": 0.0005299394292370413, | |
| "loss": 3.5311, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 5.879424778761062, | |
| "grad_norm": 0.32470381259918213, | |
| "learning_rate": 0.0005297647058823528, | |
| "loss": 3.5493, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 5.893979972054029, | |
| "grad_norm": 0.3192812204360962, | |
| "learning_rate": 0.0005295899825276645, | |
| "loss": 3.5181, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 5.908535165346996, | |
| "grad_norm": 0.3229912519454956, | |
| "learning_rate": 0.000529415259172976, | |
| "loss": 3.5267, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 5.923090358639962, | |
| "grad_norm": 0.31635522842407227, | |
| "learning_rate": 0.0005292405358182877, | |
| "loss": 3.5297, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 5.937645551932929, | |
| "grad_norm": 0.3278980851173401, | |
| "learning_rate": 0.0005290658124635992, | |
| "loss": 3.5257, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 5.9522007452258965, | |
| "grad_norm": 0.3188556134700775, | |
| "learning_rate": 0.0005288910891089109, | |
| "loss": 3.5308, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 5.9667559385188635, | |
| "grad_norm": 0.3289923071861267, | |
| "learning_rate": 0.0005287163657542224, | |
| "loss": 3.5367, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 5.981311131811831, | |
| "grad_norm": 0.2996402680873871, | |
| "learning_rate": 0.000528541642399534, | |
| "loss": 3.5295, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 5.995866325104798, | |
| "grad_norm": 0.3355785608291626, | |
| "learning_rate": 0.0005283669190448456, | |
| "loss": 3.5302, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 6.010188635305076, | |
| "grad_norm": 0.3253796696662903, | |
| "learning_rate": 0.0005281921956901572, | |
| "loss": 3.4522, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 6.0247438285980435, | |
| "grad_norm": 0.31267303228378296, | |
| "learning_rate": 0.0005280174723354688, | |
| "loss": 3.4193, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 6.0392990218910105, | |
| "grad_norm": 0.35141482949256897, | |
| "learning_rate": 0.0005278427489807804, | |
| "loss": 3.4292, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 6.053854215183978, | |
| "grad_norm": 0.31395936012268066, | |
| "learning_rate": 0.000527668025626092, | |
| "loss": 3.4291, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 6.068409408476945, | |
| "grad_norm": 0.32290589809417725, | |
| "learning_rate": 0.0005274933022714035, | |
| "loss": 3.4305, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 6.082964601769912, | |
| "grad_norm": 0.3760620951652527, | |
| "learning_rate": 0.0005273185789167152, | |
| "loss": 3.4281, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 6.097519795062879, | |
| "grad_norm": 0.3305610120296478, | |
| "learning_rate": 0.0005271438555620268, | |
| "loss": 3.4387, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 6.112074988355845, | |
| "grad_norm": 0.330484002828598, | |
| "learning_rate": 0.0005269691322073384, | |
| "loss": 3.442, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.112074988355845, | |
| "eval_accuracy": 0.3625605721041184, | |
| "eval_loss": 3.6160457134246826, | |
| "eval_runtime": 181.0216, | |
| "eval_samples_per_second": 91.984, | |
| "eval_steps_per_second": 5.751, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.126630181648812, | |
| "grad_norm": 0.3237796127796173, | |
| "learning_rate": 0.0005267944088526499, | |
| "loss": 3.4373, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 6.141185374941779, | |
| "grad_norm": 0.33345699310302734, | |
| "learning_rate": 0.0005266196854979615, | |
| "loss": 3.4453, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 6.155740568234746, | |
| "grad_norm": 0.3279930353164673, | |
| "learning_rate": 0.0005264449621432731, | |
| "loss": 3.4346, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 6.170295761527713, | |
| "grad_norm": 0.33429208397865295, | |
| "learning_rate": 0.0005262702387885847, | |
| "loss": 3.4517, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 6.18485095482068, | |
| "grad_norm": 0.32830730080604553, | |
| "learning_rate": 0.0005260955154338963, | |
| "loss": 3.4596, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 6.199406148113647, | |
| "grad_norm": 0.3260050117969513, | |
| "learning_rate": 0.0005259207920792079, | |
| "loss": 3.4487, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 6.2139613414066135, | |
| "grad_norm": 0.3051382005214691, | |
| "learning_rate": 0.0005257460687245195, | |
| "loss": 3.4669, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 6.2285165346995806, | |
| "grad_norm": 0.35948219895362854, | |
| "learning_rate": 0.000525571345369831, | |
| "loss": 3.4666, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 6.243071727992548, | |
| "grad_norm": 0.33564814925193787, | |
| "learning_rate": 0.0005253966220151426, | |
| "loss": 3.4627, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 6.257626921285515, | |
| "grad_norm": 0.31916743516921997, | |
| "learning_rate": 0.0005252218986604543, | |
| "loss": 3.4698, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 6.272182114578482, | |
| "grad_norm": 0.31790220737457275, | |
| "learning_rate": 0.0005250471753057658, | |
| "loss": 3.4645, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 6.286737307871449, | |
| "grad_norm": 0.32677316665649414, | |
| "learning_rate": 0.0005248724519510774, | |
| "loss": 3.4663, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 6.301292501164416, | |
| "grad_norm": 0.3288456201553345, | |
| "learning_rate": 0.000524697728596389, | |
| "loss": 3.4605, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 6.315847694457382, | |
| "grad_norm": 0.33529654145240784, | |
| "learning_rate": 0.0005245230052417006, | |
| "loss": 3.4611, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 6.330402887750349, | |
| "grad_norm": 0.3224530816078186, | |
| "learning_rate": 0.0005243482818870122, | |
| "loss": 3.4751, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 6.344958081043316, | |
| "grad_norm": 0.30517300963401794, | |
| "learning_rate": 0.0005241735585323238, | |
| "loss": 3.4797, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 6.359513274336283, | |
| "grad_norm": 0.31948354840278625, | |
| "learning_rate": 0.0005239988351776354, | |
| "loss": 3.4618, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 6.37406846762925, | |
| "grad_norm": 0.3166545331478119, | |
| "learning_rate": 0.0005238241118229469, | |
| "loss": 3.4755, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 6.388623660922217, | |
| "grad_norm": 0.30832865834236145, | |
| "learning_rate": 0.0005236493884682585, | |
| "loss": 3.4632, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 6.403178854215184, | |
| "grad_norm": 0.31591275334358215, | |
| "learning_rate": 0.0005234746651135701, | |
| "loss": 3.4671, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.403178854215184, | |
| "eval_accuracy": 0.36362748498265474, | |
| "eval_loss": 3.609940767288208, | |
| "eval_runtime": 180.7493, | |
| "eval_samples_per_second": 92.122, | |
| "eval_steps_per_second": 5.759, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.417734047508151, | |
| "grad_norm": 0.31404587626457214, | |
| "learning_rate": 0.0005232999417588818, | |
| "loss": 3.4756, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 6.432289240801118, | |
| "grad_norm": 0.3089274764060974, | |
| "learning_rate": 0.0005231252184041933, | |
| "loss": 3.4688, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 6.446844434094085, | |
| "grad_norm": 0.32326218485832214, | |
| "learning_rate": 0.0005229504950495049, | |
| "loss": 3.4659, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 6.461399627387052, | |
| "grad_norm": 0.3198147714138031, | |
| "learning_rate": 0.0005227757716948165, | |
| "loss": 3.4657, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 6.475954820680019, | |
| "grad_norm": 0.36542582511901855, | |
| "learning_rate": 0.000522601048340128, | |
| "loss": 3.4748, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 6.490510013972986, | |
| "grad_norm": 0.34132885932922363, | |
| "learning_rate": 0.0005224263249854397, | |
| "loss": 3.4711, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 6.505065207265952, | |
| "grad_norm": 0.3270913362503052, | |
| "learning_rate": 0.0005222516016307512, | |
| "loss": 3.4922, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 6.519620400558919, | |
| "grad_norm": 0.32891523838043213, | |
| "learning_rate": 0.0005220768782760629, | |
| "loss": 3.4697, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 6.534175593851886, | |
| "grad_norm": 0.3459501564502716, | |
| "learning_rate": 0.0005219021549213744, | |
| "loss": 3.4809, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 6.548730787144853, | |
| "grad_norm": 0.29732978343963623, | |
| "learning_rate": 0.000521727431566686, | |
| "loss": 3.4849, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 6.56328598043782, | |
| "grad_norm": 0.3077053129673004, | |
| "learning_rate": 0.0005215527082119976, | |
| "loss": 3.4712, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 6.577841173730787, | |
| "grad_norm": 0.314859002828598, | |
| "learning_rate": 0.0005213779848573093, | |
| "loss": 3.4753, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 6.592396367023754, | |
| "grad_norm": 0.35085999965667725, | |
| "learning_rate": 0.0005212032615026208, | |
| "loss": 3.4777, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 6.6069515603167215, | |
| "grad_norm": 0.33542004227638245, | |
| "learning_rate": 0.0005210285381479323, | |
| "loss": 3.4806, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 6.621506753609688, | |
| "grad_norm": 0.3448728024959564, | |
| "learning_rate": 0.000520853814793244, | |
| "loss": 3.5001, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 6.636061946902655, | |
| "grad_norm": 0.32117313146591187, | |
| "learning_rate": 0.0005206790914385555, | |
| "loss": 3.4917, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 6.650617140195622, | |
| "grad_norm": 0.33129552006721497, | |
| "learning_rate": 0.0005205043680838672, | |
| "loss": 3.484, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 6.665172333488589, | |
| "grad_norm": 0.3124433755874634, | |
| "learning_rate": 0.0005203296447291787, | |
| "loss": 3.4768, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 6.679727526781556, | |
| "grad_norm": 0.3298899829387665, | |
| "learning_rate": 0.0005201549213744904, | |
| "loss": 3.4987, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 6.694282720074523, | |
| "grad_norm": 0.32636532187461853, | |
| "learning_rate": 0.0005199801980198019, | |
| "loss": 3.4893, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.694282720074523, | |
| "eval_accuracy": 0.3642555509027801, | |
| "eval_loss": 3.598724842071533, | |
| "eval_runtime": 180.9694, | |
| "eval_samples_per_second": 92.01, | |
| "eval_steps_per_second": 5.752, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.708837913367489, | |
| "grad_norm": 0.32573258876800537, | |
| "learning_rate": 0.0005198054746651136, | |
| "loss": 3.4969, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 6.723393106660456, | |
| "grad_norm": 0.32142332196235657, | |
| "learning_rate": 0.0005196307513104251, | |
| "loss": 3.4879, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 6.737948299953423, | |
| "grad_norm": 0.3339453637599945, | |
| "learning_rate": 0.0005194560279557367, | |
| "loss": 3.4853, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 6.75250349324639, | |
| "grad_norm": 0.3392508625984192, | |
| "learning_rate": 0.0005192813046010483, | |
| "loss": 3.4777, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 6.767058686539357, | |
| "grad_norm": 0.3458227515220642, | |
| "learning_rate": 0.0005191065812463599, | |
| "loss": 3.4887, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 6.781613879832324, | |
| "grad_norm": 0.3117659389972687, | |
| "learning_rate": 0.0005189318578916715, | |
| "loss": 3.4842, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 6.7961690731252915, | |
| "grad_norm": 0.3235318958759308, | |
| "learning_rate": 0.000518757134536983, | |
| "loss": 3.4785, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 6.810724266418258, | |
| "grad_norm": 0.32054176926612854, | |
| "learning_rate": 0.0005185824111822947, | |
| "loss": 3.4813, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 6.825279459711225, | |
| "grad_norm": 0.319515585899353, | |
| "learning_rate": 0.0005184076878276063, | |
| "loss": 3.4716, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 6.839834653004192, | |
| "grad_norm": 0.3443147540092468, | |
| "learning_rate": 0.0005182329644729179, | |
| "loss": 3.4967, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 6.854389846297159, | |
| "grad_norm": 0.3200646936893463, | |
| "learning_rate": 0.0005180582411182294, | |
| "loss": 3.4893, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 6.868945039590126, | |
| "grad_norm": 0.33306822180747986, | |
| "learning_rate": 0.000517883517763541, | |
| "loss": 3.5065, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 6.883500232883093, | |
| "grad_norm": 0.3191053867340088, | |
| "learning_rate": 0.0005177087944088526, | |
| "loss": 3.5058, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 6.898055426176059, | |
| "grad_norm": 0.32728537917137146, | |
| "learning_rate": 0.0005175340710541642, | |
| "loss": 3.488, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 6.912610619469026, | |
| "grad_norm": 0.315595418214798, | |
| "learning_rate": 0.0005173593476994758, | |
| "loss": 3.4906, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 6.927165812761993, | |
| "grad_norm": 0.3276635408401489, | |
| "learning_rate": 0.0005171846243447874, | |
| "loss": 3.4941, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 6.94172100605496, | |
| "grad_norm": 0.31373319029808044, | |
| "learning_rate": 0.000517009900990099, | |
| "loss": 3.4911, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 6.956276199347927, | |
| "grad_norm": 0.33638837933540344, | |
| "learning_rate": 0.0005168351776354105, | |
| "loss": 3.4843, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 6.9708313926408945, | |
| "grad_norm": 0.33347102999687195, | |
| "learning_rate": 0.0005166604542807221, | |
| "loss": 3.4993, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 6.9853865859338615, | |
| "grad_norm": 0.3285515606403351, | |
| "learning_rate": 0.0005164857309260338, | |
| "loss": 3.4985, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.9853865859338615, | |
| "eval_accuracy": 0.3649057119563231, | |
| "eval_loss": 3.588160514831543, | |
| "eval_runtime": 180.8076, | |
| "eval_samples_per_second": 92.092, | |
| "eval_steps_per_second": 5.758, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.999941779226829, | |
| "grad_norm": 0.3126235604286194, | |
| "learning_rate": 0.0005163110075713453, | |
| "loss": 3.4946, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 7.014264089427107, | |
| "grad_norm": 0.3227352499961853, | |
| "learning_rate": 0.0005161362842166569, | |
| "loss": 3.3874, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 7.028819282720074, | |
| "grad_norm": 0.33581089973449707, | |
| "learning_rate": 0.0005159615608619685, | |
| "loss": 3.3799, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 7.0433744760130415, | |
| "grad_norm": 0.33446604013442993, | |
| "learning_rate": 0.0005157868375072801, | |
| "loss": 3.3827, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 7.0579296693060085, | |
| "grad_norm": 0.32778769731521606, | |
| "learning_rate": 0.0005156121141525917, | |
| "loss": 3.3804, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 7.072484862598976, | |
| "grad_norm": 0.3300207555294037, | |
| "learning_rate": 0.0005154373907979033, | |
| "loss": 3.3909, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 7.087040055891943, | |
| "grad_norm": 0.3466382920742035, | |
| "learning_rate": 0.0005152626674432149, | |
| "loss": 3.3882, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 7.101595249184909, | |
| "grad_norm": 0.3209831714630127, | |
| "learning_rate": 0.0005150879440885264, | |
| "loss": 3.4095, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 7.116150442477876, | |
| "grad_norm": 0.31116464734077454, | |
| "learning_rate": 0.000514913220733838, | |
| "loss": 3.405, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 7.130705635770843, | |
| "grad_norm": 0.32435309886932373, | |
| "learning_rate": 0.0005147384973791496, | |
| "loss": 3.3998, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 7.14526082906381, | |
| "grad_norm": 0.3378669321537018, | |
| "learning_rate": 0.0005145637740244613, | |
| "loss": 3.4078, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 7.159816022356777, | |
| "grad_norm": 0.3956412672996521, | |
| "learning_rate": 0.0005143890506697728, | |
| "loss": 3.4078, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 7.174371215649744, | |
| "grad_norm": 0.34623751044273376, | |
| "learning_rate": 0.0005142143273150844, | |
| "loss": 3.4175, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 7.188926408942711, | |
| "grad_norm": 0.3511153757572174, | |
| "learning_rate": 0.000514039603960396, | |
| "loss": 3.4144, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 7.203481602235677, | |
| "grad_norm": 0.30973565578460693, | |
| "learning_rate": 0.0005138648806057075, | |
| "loss": 3.4264, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 7.218036795528644, | |
| "grad_norm": 0.3381648063659668, | |
| "learning_rate": 0.0005136901572510192, | |
| "loss": 3.4011, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 7.2325919888216115, | |
| "grad_norm": 0.33070191740989685, | |
| "learning_rate": 0.0005135154338963307, | |
| "loss": 3.4212, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 7.2471471821145785, | |
| "grad_norm": 0.3330495357513428, | |
| "learning_rate": 0.0005133407105416424, | |
| "loss": 3.4229, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 7.261702375407546, | |
| "grad_norm": 0.3378702700138092, | |
| "learning_rate": 0.0005131659871869539, | |
| "loss": 3.4219, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 7.276257568700513, | |
| "grad_norm": 0.35611557960510254, | |
| "learning_rate": 0.0005129912638322656, | |
| "loss": 3.4264, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.276257568700513, | |
| "eval_accuracy": 0.36439411559586166, | |
| "eval_loss": 3.598292350769043, | |
| "eval_runtime": 180.9485, | |
| "eval_samples_per_second": 92.021, | |
| "eval_steps_per_second": 5.753, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.290812761993479, | |
| "grad_norm": 0.3345873951911926, | |
| "learning_rate": 0.0005128165404775771, | |
| "loss": 3.433, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 7.305367955286446, | |
| "grad_norm": 0.33987024426460266, | |
| "learning_rate": 0.0005126418171228888, | |
| "loss": 3.4204, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 7.319923148579413, | |
| "grad_norm": 0.3445035219192505, | |
| "learning_rate": 0.0005124670937682003, | |
| "loss": 3.4285, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 7.33447834187238, | |
| "grad_norm": 0.345106840133667, | |
| "learning_rate": 0.000512292370413512, | |
| "loss": 3.4366, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 7.349033535165347, | |
| "grad_norm": 0.3592861294746399, | |
| "learning_rate": 0.0005121176470588235, | |
| "loss": 3.4408, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 7.363588728458314, | |
| "grad_norm": 0.33465760946273804, | |
| "learning_rate": 0.000511942923704135, | |
| "loss": 3.4377, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 7.378143921751281, | |
| "grad_norm": 0.361323744058609, | |
| "learning_rate": 0.0005117682003494467, | |
| "loss": 3.4457, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 7.392699115044248, | |
| "grad_norm": 0.3065403997898102, | |
| "learning_rate": 0.0005115934769947583, | |
| "loss": 3.4445, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 7.407254308337214, | |
| "grad_norm": 0.3310743570327759, | |
| "learning_rate": 0.0005114187536400699, | |
| "loss": 3.4393, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 7.4218095016301815, | |
| "grad_norm": 0.34009453654289246, | |
| "learning_rate": 0.0005112440302853814, | |
| "loss": 3.4302, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 7.4363646949231486, | |
| "grad_norm": 0.3367420732975006, | |
| "learning_rate": 0.0005110693069306931, | |
| "loss": 3.4376, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 7.450919888216116, | |
| "grad_norm": 0.3324291706085205, | |
| "learning_rate": 0.0005108945835760046, | |
| "loss": 3.4399, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 7.465475081509083, | |
| "grad_norm": 0.32936856150627136, | |
| "learning_rate": 0.0005107198602213162, | |
| "loss": 3.4438, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 7.48003027480205, | |
| "grad_norm": 0.3215152323246002, | |
| "learning_rate": 0.0005105451368666278, | |
| "loss": 3.4416, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 7.494585468095016, | |
| "grad_norm": 0.3312242925167084, | |
| "learning_rate": 0.0005103704135119394, | |
| "loss": 3.4529, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 7.509140661387983, | |
| "grad_norm": 0.34947583079338074, | |
| "learning_rate": 0.000510195690157251, | |
| "loss": 3.4315, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 7.52369585468095, | |
| "grad_norm": 0.3578192889690399, | |
| "learning_rate": 0.0005100209668025625, | |
| "loss": 3.4518, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 7.538251047973917, | |
| "grad_norm": 0.3474888801574707, | |
| "learning_rate": 0.0005098462434478742, | |
| "loss": 3.4521, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 7.552806241266884, | |
| "grad_norm": 0.314382404088974, | |
| "learning_rate": 0.0005096715200931858, | |
| "loss": 3.4394, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 7.567361434559851, | |
| "grad_norm": 0.34033820033073425, | |
| "learning_rate": 0.0005094967967384974, | |
| "loss": 3.4448, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.567361434559851, | |
| "eval_accuracy": 0.3653323360749711, | |
| "eval_loss": 3.592108726501465, | |
| "eval_runtime": 181.0772, | |
| "eval_samples_per_second": 91.955, | |
| "eval_steps_per_second": 5.749, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.581916627852818, | |
| "grad_norm": 0.3230842053890228, | |
| "learning_rate": 0.0005093220733838089, | |
| "loss": 3.4504, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 7.5964718211457845, | |
| "grad_norm": 0.3355337977409363, | |
| "learning_rate": 0.0005091473500291205, | |
| "loss": 3.439, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 7.6110270144387515, | |
| "grad_norm": 0.34470894932746887, | |
| "learning_rate": 0.0005089726266744321, | |
| "loss": 3.4568, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 7.625582207731719, | |
| "grad_norm": 0.3162420094013214, | |
| "learning_rate": 0.0005087979033197437, | |
| "loss": 3.4457, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 7.640137401024686, | |
| "grad_norm": 0.325000524520874, | |
| "learning_rate": 0.0005086231799650553, | |
| "loss": 3.4556, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 7.654692594317653, | |
| "grad_norm": 0.32103052735328674, | |
| "learning_rate": 0.0005084484566103669, | |
| "loss": 3.4577, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 7.66924778761062, | |
| "grad_norm": 0.34083518385887146, | |
| "learning_rate": 0.0005082737332556785, | |
| "loss": 3.4528, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 7.683802980903586, | |
| "grad_norm": 0.3336086869239807, | |
| "learning_rate": 0.00050809900990099, | |
| "loss": 3.4556, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 7.698358174196553, | |
| "grad_norm": 0.3343296945095062, | |
| "learning_rate": 0.0005079242865463016, | |
| "loss": 3.4525, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 7.71291336748952, | |
| "grad_norm": 0.3320295512676239, | |
| "learning_rate": 0.0005077495631916133, | |
| "loss": 3.4512, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 7.727468560782487, | |
| "grad_norm": 0.3143779933452606, | |
| "learning_rate": 0.0005075748398369248, | |
| "loss": 3.4488, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 7.742023754075454, | |
| "grad_norm": 0.32744669914245605, | |
| "learning_rate": 0.0005074001164822364, | |
| "loss": 3.4614, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 7.756578947368421, | |
| "grad_norm": 0.3253073990345001, | |
| "learning_rate": 0.000507225393127548, | |
| "loss": 3.4551, | |
| "step": 26650 | |
| }, | |
| { | |
| "epoch": 7.771134140661388, | |
| "grad_norm": 0.3403880000114441, | |
| "learning_rate": 0.0005070506697728596, | |
| "loss": 3.4575, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 7.785689333954355, | |
| "grad_norm": 0.3548356592655182, | |
| "learning_rate": 0.0005068759464181711, | |
| "loss": 3.4624, | |
| "step": 26750 | |
| }, | |
| { | |
| "epoch": 7.8002445272473215, | |
| "grad_norm": 0.33326730132102966, | |
| "learning_rate": 0.0005067012230634828, | |
| "loss": 3.4634, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 7.814799720540289, | |
| "grad_norm": 0.35093986988067627, | |
| "learning_rate": 0.0005065264997087944, | |
| "loss": 3.4566, | |
| "step": 26850 | |
| }, | |
| { | |
| "epoch": 7.829354913833256, | |
| "grad_norm": 0.3309878706932068, | |
| "learning_rate": 0.0005063517763541059, | |
| "loss": 3.4525, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 7.843910107126223, | |
| "grad_norm": 0.33384084701538086, | |
| "learning_rate": 0.0005061770529994175, | |
| "loss": 3.4692, | |
| "step": 26950 | |
| }, | |
| { | |
| "epoch": 7.85846530041919, | |
| "grad_norm": 0.3291708827018738, | |
| "learning_rate": 0.0005060023296447291, | |
| "loss": 3.4691, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.85846530041919, | |
| "eval_accuracy": 0.36616501703382, | |
| "eval_loss": 3.5818653106689453, | |
| "eval_runtime": 181.056, | |
| "eval_samples_per_second": 91.966, | |
| "eval_steps_per_second": 5.75, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.873020493712157, | |
| "grad_norm": 0.3469105660915375, | |
| "learning_rate": 0.0005058276062900408, | |
| "loss": 3.4533, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 7.887575687005123, | |
| "grad_norm": 0.3380752205848694, | |
| "learning_rate": 0.0005056528829353523, | |
| "loss": 3.4577, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 7.90213088029809, | |
| "grad_norm": 0.3110817074775696, | |
| "learning_rate": 0.000505478159580664, | |
| "loss": 3.471, | |
| "step": 27150 | |
| }, | |
| { | |
| "epoch": 7.916686073591057, | |
| "grad_norm": 0.32800230383872986, | |
| "learning_rate": 0.0005053034362259755, | |
| "loss": 3.4548, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 7.931241266884024, | |
| "grad_norm": 0.3017429709434509, | |
| "learning_rate": 0.000505128712871287, | |
| "loss": 3.4505, | |
| "step": 27250 | |
| }, | |
| { | |
| "epoch": 7.945796460176991, | |
| "grad_norm": 0.33647236227989197, | |
| "learning_rate": 0.0005049539895165987, | |
| "loss": 3.4667, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 7.960351653469958, | |
| "grad_norm": 0.34851449728012085, | |
| "learning_rate": 0.0005047792661619103, | |
| "loss": 3.4613, | |
| "step": 27350 | |
| }, | |
| { | |
| "epoch": 7.974906846762925, | |
| "grad_norm": 0.33686619997024536, | |
| "learning_rate": 0.0005046045428072219, | |
| "loss": 3.4554, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 7.989462040055892, | |
| "grad_norm": 0.30195578932762146, | |
| "learning_rate": 0.0005044298194525334, | |
| "loss": 3.4715, | |
| "step": 27450 | |
| }, | |
| { | |
| "epoch": 8.003784350256172, | |
| "grad_norm": 0.3394927680492401, | |
| "learning_rate": 0.0005042550960978451, | |
| "loss": 3.4395, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 8.01833954354914, | |
| "grad_norm": 0.32353535294532776, | |
| "learning_rate": 0.0005040803727431566, | |
| "loss": 3.3566, | |
| "step": 27550 | |
| }, | |
| { | |
| "epoch": 8.032894736842104, | |
| "grad_norm": 0.3227969706058502, | |
| "learning_rate": 0.0005039056493884683, | |
| "loss": 3.3603, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 8.047449930135071, | |
| "grad_norm": 0.31982970237731934, | |
| "learning_rate": 0.0005037309260337798, | |
| "loss": 3.348, | |
| "step": 27650 | |
| }, | |
| { | |
| "epoch": 8.062005123428039, | |
| "grad_norm": 0.3820800483226776, | |
| "learning_rate": 0.0005035562026790914, | |
| "loss": 3.3559, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 8.076560316721006, | |
| "grad_norm": 0.3284827470779419, | |
| "learning_rate": 0.000503381479324403, | |
| "loss": 3.377, | |
| "step": 27750 | |
| }, | |
| { | |
| "epoch": 8.091115510013973, | |
| "grad_norm": 0.33183208107948303, | |
| "learning_rate": 0.0005032067559697145, | |
| "loss": 3.362, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 8.10567070330694, | |
| "grad_norm": 0.3440621495246887, | |
| "learning_rate": 0.0005030320326150262, | |
| "loss": 3.3727, | |
| "step": 27850 | |
| }, | |
| { | |
| "epoch": 8.120225896599907, | |
| "grad_norm": 0.3330867290496826, | |
| "learning_rate": 0.0005028573092603378, | |
| "loss": 3.3693, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 8.134781089892874, | |
| "grad_norm": 0.3781154155731201, | |
| "learning_rate": 0.0005026825859056494, | |
| "loss": 3.378, | |
| "step": 27950 | |
| }, | |
| { | |
| "epoch": 8.149336283185841, | |
| "grad_norm": 0.3525262773036957, | |
| "learning_rate": 0.0005025078625509609, | |
| "loss": 3.3775, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.149336283185841, | |
| "eval_accuracy": 0.36626068426042596, | |
| "eval_loss": 3.590256929397583, | |
| "eval_runtime": 180.9876, | |
| "eval_samples_per_second": 92.001, | |
| "eval_steps_per_second": 5.752, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.163891476478808, | |
| "grad_norm": 0.3500209450721741, | |
| "learning_rate": 0.0005023331391962726, | |
| "loss": 3.3713, | |
| "step": 28050 | |
| }, | |
| { | |
| "epoch": 8.178446669771775, | |
| "grad_norm": 0.32972395420074463, | |
| "learning_rate": 0.0005021584158415841, | |
| "loss": 3.3718, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 8.193001863064742, | |
| "grad_norm": 0.3366987705230713, | |
| "learning_rate": 0.0005019836924868956, | |
| "loss": 3.3944, | |
| "step": 28150 | |
| }, | |
| { | |
| "epoch": 8.20755705635771, | |
| "grad_norm": 0.32175183296203613, | |
| "learning_rate": 0.0005018089691322073, | |
| "loss": 3.374, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 8.222112249650676, | |
| "grad_norm": 0.325791597366333, | |
| "learning_rate": 0.0005016342457775189, | |
| "loss": 3.3832, | |
| "step": 28250 | |
| }, | |
| { | |
| "epoch": 8.236667442943642, | |
| "grad_norm": 0.34130793809890747, | |
| "learning_rate": 0.0005014595224228305, | |
| "loss": 3.3921, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 8.251222636236609, | |
| "grad_norm": 0.321627676486969, | |
| "learning_rate": 0.000501284799068142, | |
| "loss": 3.3985, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 8.265777829529576, | |
| "grad_norm": 0.3398359417915344, | |
| "learning_rate": 0.0005011100757134537, | |
| "loss": 3.3839, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 8.280333022822543, | |
| "grad_norm": 0.3410577178001404, | |
| "learning_rate": 0.0005009353523587653, | |
| "loss": 3.3938, | |
| "step": 28450 | |
| }, | |
| { | |
| "epoch": 8.29488821611551, | |
| "grad_norm": 0.3390825390815735, | |
| "learning_rate": 0.0005007606290040768, | |
| "loss": 3.4063, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 8.309443409408477, | |
| "grad_norm": 0.3315296471118927, | |
| "learning_rate": 0.0005005859056493884, | |
| "loss": 3.4016, | |
| "step": 28550 | |
| }, | |
| { | |
| "epoch": 8.323998602701444, | |
| "grad_norm": 0.34398385882377625, | |
| "learning_rate": 0.0005004111822947, | |
| "loss": 3.4099, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 8.338553795994411, | |
| "grad_norm": 0.3477787971496582, | |
| "learning_rate": 0.0005002364589400116, | |
| "loss": 3.4103, | |
| "step": 28650 | |
| }, | |
| { | |
| "epoch": 8.353108989287378, | |
| "grad_norm": 0.3421306908130646, | |
| "learning_rate": 0.0005000617355853231, | |
| "loss": 3.3944, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 8.367664182580345, | |
| "grad_norm": 0.33684462308883667, | |
| "learning_rate": 0.0004998870122306348, | |
| "loss": 3.409, | |
| "step": 28750 | |
| }, | |
| { | |
| "epoch": 8.382219375873312, | |
| "grad_norm": 0.314606636762619, | |
| "learning_rate": 0.0004997122888759464, | |
| "loss": 3.4149, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 8.39677456916628, | |
| "grad_norm": 0.35119929909706116, | |
| "learning_rate": 0.000499537565521258, | |
| "loss": 3.4165, | |
| "step": 28850 | |
| }, | |
| { | |
| "epoch": 8.411329762459246, | |
| "grad_norm": 0.35205650329589844, | |
| "learning_rate": 0.0004993628421665695, | |
| "loss": 3.4171, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 8.425884955752213, | |
| "grad_norm": 0.3453122079372406, | |
| "learning_rate": 0.0004991881188118811, | |
| "loss": 3.409, | |
| "step": 28950 | |
| }, | |
| { | |
| "epoch": 8.440440149045179, | |
| "grad_norm": 0.3224046230316162, | |
| "learning_rate": 0.0004990133954571928, | |
| "loss": 3.4149, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.440440149045179, | |
| "eval_accuracy": 0.3668040129933488, | |
| "eval_loss": 3.584907293319702, | |
| "eval_runtime": 180.9411, | |
| "eval_samples_per_second": 92.024, | |
| "eval_steps_per_second": 5.753, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.454995342338146, | |
| "grad_norm": 0.33683833479881287, | |
| "learning_rate": 0.0004988386721025043, | |
| "loss": 3.4179, | |
| "step": 29050 | |
| }, | |
| { | |
| "epoch": 8.469550535631113, | |
| "grad_norm": 0.31904080510139465, | |
| "learning_rate": 0.0004986639487478159, | |
| "loss": 3.4262, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 8.48410572892408, | |
| "grad_norm": 0.34945112466812134, | |
| "learning_rate": 0.0004984892253931275, | |
| "loss": 3.4071, | |
| "step": 29150 | |
| }, | |
| { | |
| "epoch": 8.498660922217047, | |
| "grad_norm": 0.3821871876716614, | |
| "learning_rate": 0.0004983145020384391, | |
| "loss": 3.412, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 8.513216115510014, | |
| "grad_norm": 0.33756762742996216, | |
| "learning_rate": 0.0004981397786837507, | |
| "loss": 3.4187, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 8.527771308802981, | |
| "grad_norm": 0.33595606684684753, | |
| "learning_rate": 0.0004979650553290622, | |
| "loss": 3.4251, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 8.542326502095948, | |
| "grad_norm": 0.3365992307662964, | |
| "learning_rate": 0.0004977903319743739, | |
| "loss": 3.4241, | |
| "step": 29350 | |
| }, | |
| { | |
| "epoch": 8.556881695388915, | |
| "grad_norm": 0.3620302677154541, | |
| "learning_rate": 0.0004976156086196854, | |
| "loss": 3.4122, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 8.571436888681882, | |
| "grad_norm": 0.3167871832847595, | |
| "learning_rate": 0.0004974408852649971, | |
| "loss": 3.4281, | |
| "step": 29450 | |
| }, | |
| { | |
| "epoch": 8.58599208197485, | |
| "grad_norm": 0.3546694815158844, | |
| "learning_rate": 0.0004972661619103086, | |
| "loss": 3.4266, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 8.600547275267816, | |
| "grad_norm": 0.33703261613845825, | |
| "learning_rate": 0.0004970914385556202, | |
| "loss": 3.4098, | |
| "step": 29550 | |
| }, | |
| { | |
| "epoch": 8.615102468560782, | |
| "grad_norm": 0.3471863567829132, | |
| "learning_rate": 0.0004969167152009318, | |
| "loss": 3.4201, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 8.629657661853749, | |
| "grad_norm": 0.34901905059814453, | |
| "learning_rate": 0.0004967419918462435, | |
| "loss": 3.4237, | |
| "step": 29650 | |
| }, | |
| { | |
| "epoch": 8.644212855146716, | |
| "grad_norm": 0.32548439502716064, | |
| "learning_rate": 0.000496567268491555, | |
| "loss": 3.4217, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 8.658768048439683, | |
| "grad_norm": 0.3385307192802429, | |
| "learning_rate": 0.0004963925451368665, | |
| "loss": 3.4222, | |
| "step": 29750 | |
| }, | |
| { | |
| "epoch": 8.67332324173265, | |
| "grad_norm": 0.33839151263237, | |
| "learning_rate": 0.0004962178217821782, | |
| "loss": 3.4217, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 8.687878435025617, | |
| "grad_norm": 0.3228484094142914, | |
| "learning_rate": 0.0004960430984274898, | |
| "loss": 3.4194, | |
| "step": 29850 | |
| }, | |
| { | |
| "epoch": 8.702433628318584, | |
| "grad_norm": 0.3400082290172577, | |
| "learning_rate": 0.0004958683750728014, | |
| "loss": 3.4276, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 8.716988821611551, | |
| "grad_norm": 0.3327467739582062, | |
| "learning_rate": 0.0004956936517181129, | |
| "loss": 3.4376, | |
| "step": 29950 | |
| }, | |
| { | |
| "epoch": 8.731544014904518, | |
| "grad_norm": 0.3484002649784088, | |
| "learning_rate": 0.0004955189283634246, | |
| "loss": 3.4252, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.731544014904518, | |
| "eval_accuracy": 0.36735321809154226, | |
| "eval_loss": 3.576219320297241, | |
| "eval_runtime": 181.0445, | |
| "eval_samples_per_second": 91.972, | |
| "eval_steps_per_second": 5.75, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.746099208197485, | |
| "grad_norm": 0.345060259103775, | |
| "learning_rate": 0.0004953442050087361, | |
| "loss": 3.4264, | |
| "step": 30050 | |
| }, | |
| { | |
| "epoch": 8.760654401490452, | |
| "grad_norm": 0.34570980072021484, | |
| "learning_rate": 0.0004951694816540476, | |
| "loss": 3.4343, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 8.77520959478342, | |
| "grad_norm": 0.357194721698761, | |
| "learning_rate": 0.0004949947582993593, | |
| "loss": 3.4181, | |
| "step": 30150 | |
| }, | |
| { | |
| "epoch": 8.789764788076386, | |
| "grad_norm": 0.3243914544582367, | |
| "learning_rate": 0.0004948200349446709, | |
| "loss": 3.4204, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 8.804319981369353, | |
| "grad_norm": 0.32040515542030334, | |
| "learning_rate": 0.0004946453115899825, | |
| "loss": 3.4296, | |
| "step": 30250 | |
| }, | |
| { | |
| "epoch": 8.81887517466232, | |
| "grad_norm": 0.34954166412353516, | |
| "learning_rate": 0.000494470588235294, | |
| "loss": 3.4327, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 8.833430367955286, | |
| "grad_norm": 0.3481607735157013, | |
| "learning_rate": 0.0004942958648806057, | |
| "loss": 3.4328, | |
| "step": 30350 | |
| }, | |
| { | |
| "epoch": 8.847985561248253, | |
| "grad_norm": 0.3422067165374756, | |
| "learning_rate": 0.0004941211415259173, | |
| "loss": 3.4195, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 8.86254075454122, | |
| "grad_norm": 0.33856475353240967, | |
| "learning_rate": 0.0004939464181712289, | |
| "loss": 3.4212, | |
| "step": 30450 | |
| }, | |
| { | |
| "epoch": 8.877095947834187, | |
| "grad_norm": 0.3440602719783783, | |
| "learning_rate": 0.0004937716948165404, | |
| "loss": 3.4384, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 8.891651141127154, | |
| "grad_norm": 0.3470291793346405, | |
| "learning_rate": 0.000493596971461852, | |
| "loss": 3.4268, | |
| "step": 30550 | |
| }, | |
| { | |
| "epoch": 8.906206334420121, | |
| "grad_norm": 0.3485357463359833, | |
| "learning_rate": 0.0004934222481071636, | |
| "loss": 3.4337, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 8.920761527713088, | |
| "grad_norm": 0.35263752937316895, | |
| "learning_rate": 0.0004932475247524751, | |
| "loss": 3.4405, | |
| "step": 30650 | |
| }, | |
| { | |
| "epoch": 8.935316721006055, | |
| "grad_norm": 0.3400113582611084, | |
| "learning_rate": 0.0004930728013977868, | |
| "loss": 3.4308, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 8.949871914299022, | |
| "grad_norm": 0.33924058079719543, | |
| "learning_rate": 0.0004928980780430984, | |
| "loss": 3.4263, | |
| "step": 30750 | |
| }, | |
| { | |
| "epoch": 8.96442710759199, | |
| "grad_norm": 0.33143338561058044, | |
| "learning_rate": 0.00049272335468841, | |
| "loss": 3.4255, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 8.978982300884956, | |
| "grad_norm": 0.334474116563797, | |
| "learning_rate": 0.0004925486313337215, | |
| "loss": 3.4441, | |
| "step": 30850 | |
| }, | |
| { | |
| "epoch": 8.993537494177923, | |
| "grad_norm": 0.3472048044204712, | |
| "learning_rate": 0.0004923739079790332, | |
| "loss": 3.4464, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 9.007859804378203, | |
| "grad_norm": 0.3369418680667877, | |
| "learning_rate": 0.0004921991846243447, | |
| "loss": 3.3817, | |
| "step": 30950 | |
| }, | |
| { | |
| "epoch": 9.022414997671168, | |
| "grad_norm": 0.36469414830207825, | |
| "learning_rate": 0.0004920244612696563, | |
| "loss": 3.3288, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.022414997671168, | |
| "eval_accuracy": 0.3673842453001712, | |
| "eval_loss": 3.5799498558044434, | |
| "eval_runtime": 181.0711, | |
| "eval_samples_per_second": 91.958, | |
| "eval_steps_per_second": 5.749, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.036970190964135, | |
| "grad_norm": 0.32988861203193665, | |
| "learning_rate": 0.0004918497379149679, | |
| "loss": 3.3293, | |
| "step": 31050 | |
| }, | |
| { | |
| "epoch": 9.051525384257102, | |
| "grad_norm": 0.3383880853652954, | |
| "learning_rate": 0.0004916750145602795, | |
| "loss": 3.3384, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 9.06608057755007, | |
| "grad_norm": 0.32971954345703125, | |
| "learning_rate": 0.0004915002912055911, | |
| "loss": 3.3286, | |
| "step": 31150 | |
| }, | |
| { | |
| "epoch": 9.080635770843037, | |
| "grad_norm": 0.3277758061885834, | |
| "learning_rate": 0.0004913255678509026, | |
| "loss": 3.3446, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 9.095190964136004, | |
| "grad_norm": 0.36317312717437744, | |
| "learning_rate": 0.0004911508444962143, | |
| "loss": 3.3389, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 9.10974615742897, | |
| "grad_norm": 0.368633896112442, | |
| "learning_rate": 0.0004909761211415259, | |
| "loss": 3.347, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 9.124301350721938, | |
| "grad_norm": 0.3367079198360443, | |
| "learning_rate": 0.0004908013977868375, | |
| "loss": 3.3551, | |
| "step": 31350 | |
| }, | |
| { | |
| "epoch": 9.138856544014905, | |
| "grad_norm": 0.36006441712379456, | |
| "learning_rate": 0.0004906266744321491, | |
| "loss": 3.3436, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 9.153411737307872, | |
| "grad_norm": 0.3398313820362091, | |
| "learning_rate": 0.0004904519510774606, | |
| "loss": 3.3646, | |
| "step": 31450 | |
| }, | |
| { | |
| "epoch": 9.167966930600839, | |
| "grad_norm": 0.3552166819572449, | |
| "learning_rate": 0.0004902772277227722, | |
| "loss": 3.351, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 9.182522123893806, | |
| "grad_norm": 0.3333413600921631, | |
| "learning_rate": 0.0004901025043680838, | |
| "loss": 3.358, | |
| "step": 31550 | |
| }, | |
| { | |
| "epoch": 9.197077317186773, | |
| "grad_norm": 0.3085155189037323, | |
| "learning_rate": 0.0004899277810133955, | |
| "loss": 3.3575, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 9.211632510479738, | |
| "grad_norm": 0.3518926501274109, | |
| "learning_rate": 0.000489753057658707, | |
| "loss": 3.3689, | |
| "step": 31650 | |
| }, | |
| { | |
| "epoch": 9.226187703772705, | |
| "grad_norm": 0.3818323612213135, | |
| "learning_rate": 0.0004895783343040186, | |
| "loss": 3.3583, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 9.240742897065672, | |
| "grad_norm": 0.3476870357990265, | |
| "learning_rate": 0.0004894036109493302, | |
| "loss": 3.3651, | |
| "step": 31750 | |
| }, | |
| { | |
| "epoch": 9.25529809035864, | |
| "grad_norm": 0.3409420847892761, | |
| "learning_rate": 0.0004892288875946419, | |
| "loss": 3.3707, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 9.269853283651607, | |
| "grad_norm": 0.32737651467323303, | |
| "learning_rate": 0.0004890541642399534, | |
| "loss": 3.3694, | |
| "step": 31850 | |
| }, | |
| { | |
| "epoch": 9.284408476944574, | |
| "grad_norm": 0.3541163206100464, | |
| "learning_rate": 0.0004888794408852649, | |
| "loss": 3.3713, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 9.29896367023754, | |
| "grad_norm": 0.3474277853965759, | |
| "learning_rate": 0.0004887047175305766, | |
| "loss": 3.3778, | |
| "step": 31950 | |
| }, | |
| { | |
| "epoch": 9.313518863530508, | |
| "grad_norm": 0.32477056980133057, | |
| "learning_rate": 0.0004885299941758881, | |
| "loss": 3.3776, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.313518863530508, | |
| "eval_accuracy": 0.36725214460888733, | |
| "eval_loss": 3.5803961753845215, | |
| "eval_runtime": 181.1965, | |
| "eval_samples_per_second": 91.895, | |
| "eval_steps_per_second": 5.745, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.328074056823475, | |
| "grad_norm": 0.3401997685432434, | |
| "learning_rate": 0.0004883552708211997, | |
| "loss": 3.3788, | |
| "step": 32050 | |
| }, | |
| { | |
| "epoch": 9.342629250116442, | |
| "grad_norm": 0.3674589693546295, | |
| "learning_rate": 0.00048818054746651137, | |
| "loss": 3.3687, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 9.357184443409409, | |
| "grad_norm": 0.3335835337638855, | |
| "learning_rate": 0.0004880058241118229, | |
| "loss": 3.3841, | |
| "step": 32150 | |
| }, | |
| { | |
| "epoch": 9.371739636702376, | |
| "grad_norm": 0.3415198028087616, | |
| "learning_rate": 0.0004878311007571345, | |
| "loss": 3.3814, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 9.386294829995343, | |
| "grad_norm": 0.3449629247188568, | |
| "learning_rate": 0.0004876563774024461, | |
| "loss": 3.3807, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 9.40085002328831, | |
| "grad_norm": 0.3588092625141144, | |
| "learning_rate": 0.00048748165404775763, | |
| "loss": 3.38, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 9.415405216581275, | |
| "grad_norm": 0.3250521421432495, | |
| "learning_rate": 0.0004873069306930693, | |
| "loss": 3.3881, | |
| "step": 32350 | |
| }, | |
| { | |
| "epoch": 9.429960409874242, | |
| "grad_norm": 0.34006109833717346, | |
| "learning_rate": 0.0004871322073383809, | |
| "loss": 3.3951, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 9.44451560316721, | |
| "grad_norm": 0.34316784143447876, | |
| "learning_rate": 0.00048695748398369247, | |
| "loss": 3.3704, | |
| "step": 32450 | |
| }, | |
| { | |
| "epoch": 9.459070796460177, | |
| "grad_norm": 0.34885114431381226, | |
| "learning_rate": 0.000486782760629004, | |
| "loss": 3.3855, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 9.473625989753144, | |
| "grad_norm": 0.36461853981018066, | |
| "learning_rate": 0.0004866080372743156, | |
| "loss": 3.3787, | |
| "step": 32550 | |
| }, | |
| { | |
| "epoch": 9.48818118304611, | |
| "grad_norm": 0.3631289005279541, | |
| "learning_rate": 0.0004864333139196272, | |
| "loss": 3.3854, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 9.502736376339078, | |
| "grad_norm": 0.3386242985725403, | |
| "learning_rate": 0.00048625859056493885, | |
| "loss": 3.382, | |
| "step": 32650 | |
| }, | |
| { | |
| "epoch": 9.517291569632045, | |
| "grad_norm": 0.3215729594230652, | |
| "learning_rate": 0.0004860838672102504, | |
| "loss": 3.3778, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 9.531846762925012, | |
| "grad_norm": 0.32870006561279297, | |
| "learning_rate": 0.000485909143855562, | |
| "loss": 3.394, | |
| "step": 32750 | |
| }, | |
| { | |
| "epoch": 9.546401956217979, | |
| "grad_norm": 0.34010738134384155, | |
| "learning_rate": 0.0004857344205008736, | |
| "loss": 3.393, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 9.560957149510946, | |
| "grad_norm": 0.32582029700279236, | |
| "learning_rate": 0.00048555969714618517, | |
| "loss": 3.3919, | |
| "step": 32850 | |
| }, | |
| { | |
| "epoch": 9.575512342803913, | |
| "grad_norm": 0.3527780771255493, | |
| "learning_rate": 0.0004853849737914967, | |
| "loss": 3.4016, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 9.59006753609688, | |
| "grad_norm": 0.3426138162612915, | |
| "learning_rate": 0.00048521025043680836, | |
| "loss": 3.3947, | |
| "step": 32950 | |
| }, | |
| { | |
| "epoch": 9.604622729389845, | |
| "grad_norm": 0.3326384425163269, | |
| "learning_rate": 0.00048503552708211995, | |
| "loss": 3.3921, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.604622729389845, | |
| "eval_accuracy": 0.36779582592372645, | |
| "eval_loss": 3.571920394897461, | |
| "eval_runtime": 180.8342, | |
| "eval_samples_per_second": 92.079, | |
| "eval_steps_per_second": 5.757, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.619177922682812, | |
| "grad_norm": 0.3384208679199219, | |
| "learning_rate": 0.00048486080372743155, | |
| "loss": 3.4002, | |
| "step": 33050 | |
| }, | |
| { | |
| "epoch": 9.63373311597578, | |
| "grad_norm": 0.3510007858276367, | |
| "learning_rate": 0.0004846860803727431, | |
| "loss": 3.403, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 9.648288309268747, | |
| "grad_norm": 0.35957035422325134, | |
| "learning_rate": 0.0004845113570180547, | |
| "loss": 3.3999, | |
| "step": 33150 | |
| }, | |
| { | |
| "epoch": 9.662843502561714, | |
| "grad_norm": 0.3399753272533417, | |
| "learning_rate": 0.00048433663366336633, | |
| "loss": 3.3905, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 9.67739869585468, | |
| "grad_norm": 0.3327416479587555, | |
| "learning_rate": 0.0004841619103086779, | |
| "loss": 3.4005, | |
| "step": 33250 | |
| }, | |
| { | |
| "epoch": 9.691953889147648, | |
| "grad_norm": 0.35132935643196106, | |
| "learning_rate": 0.00048398718695398947, | |
| "loss": 3.4131, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 9.706509082440615, | |
| "grad_norm": 0.3293425440788269, | |
| "learning_rate": 0.00048381246359930106, | |
| "loss": 3.3982, | |
| "step": 33350 | |
| }, | |
| { | |
| "epoch": 9.721064275733582, | |
| "grad_norm": 0.3519092798233032, | |
| "learning_rate": 0.00048363774024461265, | |
| "loss": 3.4061, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 9.735619469026549, | |
| "grad_norm": 0.3286028802394867, | |
| "learning_rate": 0.0004834630168899242, | |
| "loss": 3.3979, | |
| "step": 33450 | |
| }, | |
| { | |
| "epoch": 9.750174662319516, | |
| "grad_norm": 0.3378731310367584, | |
| "learning_rate": 0.00048328829353523584, | |
| "loss": 3.388, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 9.764729855612483, | |
| "grad_norm": 0.3523999750614166, | |
| "learning_rate": 0.00048311357018054744, | |
| "loss": 3.4064, | |
| "step": 33550 | |
| }, | |
| { | |
| "epoch": 9.77928504890545, | |
| "grad_norm": 0.3340063989162445, | |
| "learning_rate": 0.00048293884682585903, | |
| "loss": 3.41, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 9.793840242198417, | |
| "grad_norm": 0.3363160490989685, | |
| "learning_rate": 0.00048276412347117057, | |
| "loss": 3.4068, | |
| "step": 33650 | |
| }, | |
| { | |
| "epoch": 9.808395435491384, | |
| "grad_norm": 0.3576688766479492, | |
| "learning_rate": 0.00048258940011648217, | |
| "loss": 3.409, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 9.82295062878435, | |
| "grad_norm": 0.34224554896354675, | |
| "learning_rate": 0.0004824146767617938, | |
| "loss": 3.4088, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 9.837505822077317, | |
| "grad_norm": 0.3352258801460266, | |
| "learning_rate": 0.0004822399534071054, | |
| "loss": 3.4147, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 9.852061015370284, | |
| "grad_norm": 0.32171106338500977, | |
| "learning_rate": 0.00048206523005241695, | |
| "loss": 3.4071, | |
| "step": 33850 | |
| }, | |
| { | |
| "epoch": 9.86661620866325, | |
| "grad_norm": 0.32649844884872437, | |
| "learning_rate": 0.00048189050669772854, | |
| "loss": 3.4206, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 9.881171401956218, | |
| "grad_norm": 0.3591761589050293, | |
| "learning_rate": 0.00048171578334304014, | |
| "loss": 3.405, | |
| "step": 33950 | |
| }, | |
| { | |
| "epoch": 9.895726595249185, | |
| "grad_norm": 0.34346505999565125, | |
| "learning_rate": 0.00048154105998835173, | |
| "loss": 3.4212, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.895726595249185, | |
| "eval_accuracy": 0.368610760259458, | |
| "eval_loss": 3.5636332035064697, | |
| "eval_runtime": 180.8038, | |
| "eval_samples_per_second": 92.094, | |
| "eval_steps_per_second": 5.758, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.910281788542152, | |
| "grad_norm": 0.33863702416419983, | |
| "learning_rate": 0.0004813663366336633, | |
| "loss": 3.41, | |
| "step": 34050 | |
| }, | |
| { | |
| "epoch": 9.924836981835119, | |
| "grad_norm": 0.3576464056968689, | |
| "learning_rate": 0.0004811916132789749, | |
| "loss": 3.4112, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 9.939392175128086, | |
| "grad_norm": 0.35173270106315613, | |
| "learning_rate": 0.0004810168899242865, | |
| "loss": 3.4118, | |
| "step": 34150 | |
| }, | |
| { | |
| "epoch": 9.953947368421053, | |
| "grad_norm": 0.3575360178947449, | |
| "learning_rate": 0.0004808421665695981, | |
| "loss": 3.4144, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 9.96850256171402, | |
| "grad_norm": 0.34621626138687134, | |
| "learning_rate": 0.00048066744321490965, | |
| "loss": 3.4117, | |
| "step": 34250 | |
| }, | |
| { | |
| "epoch": 9.983057755006987, | |
| "grad_norm": 0.3460022211074829, | |
| "learning_rate": 0.00048049271986022124, | |
| "loss": 3.4004, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 9.997612948299953, | |
| "grad_norm": 0.3510585427284241, | |
| "learning_rate": 0.0004803179965055329, | |
| "loss": 3.4003, | |
| "step": 34350 | |
| }, | |
| { | |
| "epoch": 10.011935258500232, | |
| "grad_norm": 0.3488433361053467, | |
| "learning_rate": 0.0004801432731508445, | |
| "loss": 3.3164, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 10.0264904517932, | |
| "grad_norm": 0.34920355677604675, | |
| "learning_rate": 0.000479968549796156, | |
| "loss": 3.2976, | |
| "step": 34450 | |
| }, | |
| { | |
| "epoch": 10.041045645086166, | |
| "grad_norm": 0.33996644616127014, | |
| "learning_rate": 0.0004797938264414676, | |
| "loss": 3.3053, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 10.055600838379133, | |
| "grad_norm": 0.3447495102882385, | |
| "learning_rate": 0.0004796191030867792, | |
| "loss": 3.3029, | |
| "step": 34550 | |
| }, | |
| { | |
| "epoch": 10.0701560316721, | |
| "grad_norm": 0.3605789244174957, | |
| "learning_rate": 0.00047944437973209086, | |
| "loss": 3.3058, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 10.084711224965067, | |
| "grad_norm": 0.34010049700737, | |
| "learning_rate": 0.0004792696563774024, | |
| "loss": 3.3146, | |
| "step": 34650 | |
| }, | |
| { | |
| "epoch": 10.099266418258035, | |
| "grad_norm": 0.35634300112724304, | |
| "learning_rate": 0.000479094933022714, | |
| "loss": 3.3178, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 10.113821611551002, | |
| "grad_norm": 0.3321462571620941, | |
| "learning_rate": 0.0004789202096680256, | |
| "loss": 3.2983, | |
| "step": 34750 | |
| }, | |
| { | |
| "epoch": 10.128376804843969, | |
| "grad_norm": 0.3477012813091278, | |
| "learning_rate": 0.00047874548631333713, | |
| "loss": 3.3279, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 10.142931998136936, | |
| "grad_norm": 0.3322581946849823, | |
| "learning_rate": 0.0004785707629586487, | |
| "loss": 3.3356, | |
| "step": 34850 | |
| }, | |
| { | |
| "epoch": 10.157487191429903, | |
| "grad_norm": 0.3342720866203308, | |
| "learning_rate": 0.0004783960396039604, | |
| "loss": 3.3184, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 10.17204238472287, | |
| "grad_norm": 0.35528644919395447, | |
| "learning_rate": 0.00047822131624927197, | |
| "loss": 3.3323, | |
| "step": 34950 | |
| }, | |
| { | |
| "epoch": 10.186597578015837, | |
| "grad_norm": 0.32312092185020447, | |
| "learning_rate": 0.0004780465928945835, | |
| "loss": 3.3329, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.186597578015837, | |
| "eval_accuracy": 0.3682773352940022, | |
| "eval_loss": 3.570481300354004, | |
| "eval_runtime": 181.1752, | |
| "eval_samples_per_second": 91.905, | |
| "eval_steps_per_second": 5.746, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.201152771308802, | |
| "grad_norm": 0.3692251443862915, | |
| "learning_rate": 0.0004778718695398951, | |
| "loss": 3.358, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 10.21570796460177, | |
| "grad_norm": 0.3326742351055145, | |
| "learning_rate": 0.0004776971461852067, | |
| "loss": 3.3378, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 10.230263157894736, | |
| "grad_norm": 0.3492138683795929, | |
| "learning_rate": 0.00047752242283051835, | |
| "loss": 3.335, | |
| "step": 35150 | |
| }, | |
| { | |
| "epoch": 10.244818351187703, | |
| "grad_norm": 0.34393349289894104, | |
| "learning_rate": 0.00047734769947582994, | |
| "loss": 3.3608, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 10.25937354448067, | |
| "grad_norm": 0.3394237458705902, | |
| "learning_rate": 0.0004771729761211415, | |
| "loss": 3.3426, | |
| "step": 35250 | |
| }, | |
| { | |
| "epoch": 10.273928737773637, | |
| "grad_norm": 0.352071613073349, | |
| "learning_rate": 0.0004769982527664531, | |
| "loss": 3.3391, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 10.288483931066605, | |
| "grad_norm": 0.35096055269241333, | |
| "learning_rate": 0.00047682352941176467, | |
| "loss": 3.3605, | |
| "step": 35350 | |
| }, | |
| { | |
| "epoch": 10.303039124359572, | |
| "grad_norm": 0.368264377117157, | |
| "learning_rate": 0.0004766488060570762, | |
| "loss": 3.3471, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 10.317594317652539, | |
| "grad_norm": 0.3177869915962219, | |
| "learning_rate": 0.00047647408270238786, | |
| "loss": 3.3564, | |
| "step": 35450 | |
| }, | |
| { | |
| "epoch": 10.332149510945506, | |
| "grad_norm": 0.36875662207603455, | |
| "learning_rate": 0.00047629935934769945, | |
| "loss": 3.3499, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 10.346704704238473, | |
| "grad_norm": 0.34088993072509766, | |
| "learning_rate": 0.00047612463599301105, | |
| "loss": 3.3666, | |
| "step": 35550 | |
| }, | |
| { | |
| "epoch": 10.36125989753144, | |
| "grad_norm": 0.35179397463798523, | |
| "learning_rate": 0.0004759499126383226, | |
| "loss": 3.3562, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 10.375815090824407, | |
| "grad_norm": 0.3595273792743683, | |
| "learning_rate": 0.0004757751892836342, | |
| "loss": 3.36, | |
| "step": 35650 | |
| }, | |
| { | |
| "epoch": 10.390370284117374, | |
| "grad_norm": 0.3881017565727234, | |
| "learning_rate": 0.0004756004659289458, | |
| "loss": 3.3608, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 10.40492547741034, | |
| "grad_norm": 0.3473113179206848, | |
| "learning_rate": 0.0004754257425742574, | |
| "loss": 3.3539, | |
| "step": 35750 | |
| }, | |
| { | |
| "epoch": 10.419480670703306, | |
| "grad_norm": 0.331753134727478, | |
| "learning_rate": 0.00047525101921956896, | |
| "loss": 3.3595, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 10.434035863996273, | |
| "grad_norm": 0.3324008882045746, | |
| "learning_rate": 0.00047507629586488056, | |
| "loss": 3.3727, | |
| "step": 35850 | |
| }, | |
| { | |
| "epoch": 10.44859105728924, | |
| "grad_norm": 0.3285880982875824, | |
| "learning_rate": 0.00047490157251019215, | |
| "loss": 3.3584, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 10.463146250582207, | |
| "grad_norm": 0.35034114122390747, | |
| "learning_rate": 0.0004747268491555037, | |
| "loss": 3.3613, | |
| "step": 35950 | |
| }, | |
| { | |
| "epoch": 10.477701443875175, | |
| "grad_norm": 0.3423924148082733, | |
| "learning_rate": 0.00047455212580081534, | |
| "loss": 3.367, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.477701443875175, | |
| "eval_accuracy": 0.3687240565818758, | |
| "eval_loss": 3.565606117248535, | |
| "eval_runtime": 181.094, | |
| "eval_samples_per_second": 91.947, | |
| "eval_steps_per_second": 5.748, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.492256637168142, | |
| "grad_norm": 0.342632532119751, | |
| "learning_rate": 0.00047437740244612694, | |
| "loss": 3.3662, | |
| "step": 36050 | |
| }, | |
| { | |
| "epoch": 10.506811830461109, | |
| "grad_norm": 0.3601525127887726, | |
| "learning_rate": 0.00047420267909143853, | |
| "loss": 3.3598, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 10.521367023754076, | |
| "grad_norm": 0.3272550702095032, | |
| "learning_rate": 0.0004740279557367501, | |
| "loss": 3.3775, | |
| "step": 36150 | |
| }, | |
| { | |
| "epoch": 10.535922217047043, | |
| "grad_norm": 0.3502359688282013, | |
| "learning_rate": 0.00047385323238206166, | |
| "loss": 3.3747, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 10.55047741034001, | |
| "grad_norm": 0.34133386611938477, | |
| "learning_rate": 0.00047367850902737326, | |
| "loss": 3.3727, | |
| "step": 36250 | |
| }, | |
| { | |
| "epoch": 10.565032603632977, | |
| "grad_norm": 0.3748016953468323, | |
| "learning_rate": 0.0004735037856726849, | |
| "loss": 3.378, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 10.579587796925944, | |
| "grad_norm": 0.3498660922050476, | |
| "learning_rate": 0.0004733290623179965, | |
| "loss": 3.3736, | |
| "step": 36350 | |
| }, | |
| { | |
| "epoch": 10.59414299021891, | |
| "grad_norm": 0.33588114380836487, | |
| "learning_rate": 0.00047315433896330804, | |
| "loss": 3.373, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 10.608698183511876, | |
| "grad_norm": 0.3481428325176239, | |
| "learning_rate": 0.00047297961560861964, | |
| "loss": 3.3796, | |
| "step": 36450 | |
| }, | |
| { | |
| "epoch": 10.623253376804843, | |
| "grad_norm": 0.3276902437210083, | |
| "learning_rate": 0.00047280489225393123, | |
| "loss": 3.3677, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 10.63780857009781, | |
| "grad_norm": 0.3404669165611267, | |
| "learning_rate": 0.0004726301688992429, | |
| "loss": 3.3745, | |
| "step": 36550 | |
| }, | |
| { | |
| "epoch": 10.652363763390778, | |
| "grad_norm": 0.34793999791145325, | |
| "learning_rate": 0.0004724554455445544, | |
| "loss": 3.3816, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 10.666918956683745, | |
| "grad_norm": 0.32581639289855957, | |
| "learning_rate": 0.000472280722189866, | |
| "loss": 3.3818, | |
| "step": 36650 | |
| }, | |
| { | |
| "epoch": 10.681474149976712, | |
| "grad_norm": 0.3253845274448395, | |
| "learning_rate": 0.0004721059988351776, | |
| "loss": 3.376, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 10.696029343269679, | |
| "grad_norm": 0.341382771730423, | |
| "learning_rate": 0.00047193127548048915, | |
| "loss": 3.3828, | |
| "step": 36750 | |
| }, | |
| { | |
| "epoch": 10.710584536562646, | |
| "grad_norm": 0.36883407831192017, | |
| "learning_rate": 0.00047175655212580074, | |
| "loss": 3.3844, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 10.725139729855613, | |
| "grad_norm": 0.3668972849845886, | |
| "learning_rate": 0.0004715818287711124, | |
| "loss": 3.3904, | |
| "step": 36850 | |
| }, | |
| { | |
| "epoch": 10.73969492314858, | |
| "grad_norm": 0.35706183314323425, | |
| "learning_rate": 0.000471407105416424, | |
| "loss": 3.3934, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 10.754250116441547, | |
| "grad_norm": 0.33550548553466797, | |
| "learning_rate": 0.0004712323820617355, | |
| "loss": 3.3862, | |
| "step": 36950 | |
| }, | |
| { | |
| "epoch": 10.768805309734514, | |
| "grad_norm": 0.3570643961429596, | |
| "learning_rate": 0.0004710576587070471, | |
| "loss": 3.3889, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.768805309734514, | |
| "eval_accuracy": 0.3696497016393061, | |
| "eval_loss": 3.5564565658569336, | |
| "eval_runtime": 181.0728, | |
| "eval_samples_per_second": 91.958, | |
| "eval_steps_per_second": 5.749, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.783360503027481, | |
| "grad_norm": 0.3547266125679016, | |
| "learning_rate": 0.0004708829353523587, | |
| "loss": 3.3802, | |
| "step": 37050 | |
| }, | |
| { | |
| "epoch": 10.797915696320446, | |
| "grad_norm": 0.3460685908794403, | |
| "learning_rate": 0.0004707082119976703, | |
| "loss": 3.3797, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 10.812470889613413, | |
| "grad_norm": 0.32947590947151184, | |
| "learning_rate": 0.0004705334886429819, | |
| "loss": 3.3932, | |
| "step": 37150 | |
| }, | |
| { | |
| "epoch": 10.82702608290638, | |
| "grad_norm": 0.33554786443710327, | |
| "learning_rate": 0.0004703587652882935, | |
| "loss": 3.3781, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 10.841581276199348, | |
| "grad_norm": 0.33499568700790405, | |
| "learning_rate": 0.0004701840419336051, | |
| "loss": 3.3761, | |
| "step": 37250 | |
| }, | |
| { | |
| "epoch": 10.856136469492315, | |
| "grad_norm": 0.3423595428466797, | |
| "learning_rate": 0.0004700093185789167, | |
| "loss": 3.3915, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 10.870691662785282, | |
| "grad_norm": 0.33275771141052246, | |
| "learning_rate": 0.0004698345952242282, | |
| "loss": 3.3884, | |
| "step": 37350 | |
| }, | |
| { | |
| "epoch": 10.885246856078249, | |
| "grad_norm": 0.3475160300731659, | |
| "learning_rate": 0.00046965987186953987, | |
| "loss": 3.3939, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 10.899802049371216, | |
| "grad_norm": 0.34615978598594666, | |
| "learning_rate": 0.00046948514851485147, | |
| "loss": 3.3877, | |
| "step": 37450 | |
| }, | |
| { | |
| "epoch": 10.914357242664183, | |
| "grad_norm": 0.3393753468990326, | |
| "learning_rate": 0.00046931042516016306, | |
| "loss": 3.3759, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 10.92891243595715, | |
| "grad_norm": 0.36012470722198486, | |
| "learning_rate": 0.0004691357018054746, | |
| "loss": 3.3791, | |
| "step": 37550 | |
| }, | |
| { | |
| "epoch": 10.943467629250117, | |
| "grad_norm": 0.33901482820510864, | |
| "learning_rate": 0.0004689609784507862, | |
| "loss": 3.3967, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 10.958022822543084, | |
| "grad_norm": 0.32990455627441406, | |
| "learning_rate": 0.0004687862550960978, | |
| "loss": 3.4031, | |
| "step": 37650 | |
| }, | |
| { | |
| "epoch": 10.972578015836051, | |
| "grad_norm": 0.379361093044281, | |
| "learning_rate": 0.00046861153174140944, | |
| "loss": 3.3777, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 10.987133209129016, | |
| "grad_norm": 0.33749642968177795, | |
| "learning_rate": 0.000468436808386721, | |
| "loss": 3.3842, | |
| "step": 37750 | |
| }, | |
| { | |
| "epoch": 11.001455519329296, | |
| "grad_norm": 0.37495532631874084, | |
| "learning_rate": 0.0004682620850320326, | |
| "loss": 3.3758, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 11.016010712622263, | |
| "grad_norm": 0.3221411406993866, | |
| "learning_rate": 0.00046808736167734417, | |
| "loss": 3.2876, | |
| "step": 37850 | |
| }, | |
| { | |
| "epoch": 11.03056590591523, | |
| "grad_norm": 0.34512948989868164, | |
| "learning_rate": 0.0004679126383226557, | |
| "loss": 3.2815, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 11.045121099208197, | |
| "grad_norm": 0.3403730094432831, | |
| "learning_rate": 0.00046773791496796736, | |
| "loss": 3.3007, | |
| "step": 37950 | |
| }, | |
| { | |
| "epoch": 11.059676292501164, | |
| "grad_norm": 0.33120304346084595, | |
| "learning_rate": 0.00046756319161327895, | |
| "loss": 3.2762, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.059676292501164, | |
| "eval_accuracy": 0.3692904206666595, | |
| "eval_loss": 3.568734645843506, | |
| "eval_runtime": 181.2074, | |
| "eval_samples_per_second": 91.889, | |
| "eval_steps_per_second": 5.745, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.074231485794131, | |
| "grad_norm": 0.3604724407196045, | |
| "learning_rate": 0.00046738846825859054, | |
| "loss": 3.2975, | |
| "step": 38050 | |
| }, | |
| { | |
| "epoch": 11.088786679087098, | |
| "grad_norm": 0.3622620403766632, | |
| "learning_rate": 0.0004672137449039021, | |
| "loss": 3.2926, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 11.103341872380065, | |
| "grad_norm": 0.35401245951652527, | |
| "learning_rate": 0.0004670390215492137, | |
| "loss": 3.2959, | |
| "step": 38150 | |
| }, | |
| { | |
| "epoch": 11.117897065673032, | |
| "grad_norm": 0.3392096161842346, | |
| "learning_rate": 0.0004668642981945253, | |
| "loss": 3.2822, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 11.132452258966, | |
| "grad_norm": 0.33189859986305237, | |
| "learning_rate": 0.0004666895748398369, | |
| "loss": 3.289, | |
| "step": 38250 | |
| }, | |
| { | |
| "epoch": 11.147007452258967, | |
| "grad_norm": 0.3406781554222107, | |
| "learning_rate": 0.00046651485148514846, | |
| "loss": 3.3034, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 11.161562645551934, | |
| "grad_norm": 0.34871330857276917, | |
| "learning_rate": 0.00046634012813046006, | |
| "loss": 3.3274, | |
| "step": 38350 | |
| }, | |
| { | |
| "epoch": 11.1761178388449, | |
| "grad_norm": 0.36096620559692383, | |
| "learning_rate": 0.00046616540477577165, | |
| "loss": 3.3217, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 11.190673032137866, | |
| "grad_norm": 0.3496626913547516, | |
| "learning_rate": 0.00046599068142108324, | |
| "loss": 3.3182, | |
| "step": 38450 | |
| }, | |
| { | |
| "epoch": 11.205228225430833, | |
| "grad_norm": 0.3420194089412689, | |
| "learning_rate": 0.0004658159580663948, | |
| "loss": 3.3136, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 11.2197834187238, | |
| "grad_norm": 0.3360060155391693, | |
| "learning_rate": 0.00046564123471170643, | |
| "loss": 3.3192, | |
| "step": 38550 | |
| }, | |
| { | |
| "epoch": 11.234338612016767, | |
| "grad_norm": 0.3573230803012848, | |
| "learning_rate": 0.00046546651135701803, | |
| "loss": 3.3189, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 11.248893805309734, | |
| "grad_norm": 0.3675867021083832, | |
| "learning_rate": 0.0004652917880023296, | |
| "loss": 3.319, | |
| "step": 38650 | |
| }, | |
| { | |
| "epoch": 11.263448998602701, | |
| "grad_norm": 0.3547263443470001, | |
| "learning_rate": 0.00046511706464764116, | |
| "loss": 3.3308, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 11.278004191895668, | |
| "grad_norm": 0.34671658277511597, | |
| "learning_rate": 0.00046494234129295276, | |
| "loss": 3.329, | |
| "step": 38750 | |
| }, | |
| { | |
| "epoch": 11.292559385188635, | |
| "grad_norm": 0.35404467582702637, | |
| "learning_rate": 0.0004647676179382644, | |
| "loss": 3.3238, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 11.307114578481603, | |
| "grad_norm": 0.3403128385543823, | |
| "learning_rate": 0.000464592894583576, | |
| "loss": 3.3302, | |
| "step": 38850 | |
| }, | |
| { | |
| "epoch": 11.32166977177457, | |
| "grad_norm": 0.3744731843471527, | |
| "learning_rate": 0.00046441817122888754, | |
| "loss": 3.3324, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 11.336224965067537, | |
| "grad_norm": 0.3594568073749542, | |
| "learning_rate": 0.00046424344787419913, | |
| "loss": 3.343, | |
| "step": 38950 | |
| }, | |
| { | |
| "epoch": 11.350780158360504, | |
| "grad_norm": 0.3566104769706726, | |
| "learning_rate": 0.00046406872451951073, | |
| "loss": 3.3424, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.350780158360504, | |
| "eval_accuracy": 0.3692883051751621, | |
| "eval_loss": 3.566476345062256, | |
| "eval_runtime": 181.1872, | |
| "eval_samples_per_second": 91.899, | |
| "eval_steps_per_second": 5.745, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.36533535165347, | |
| "grad_norm": 0.3528974950313568, | |
| "learning_rate": 0.00046389400116482227, | |
| "loss": 3.3309, | |
| "step": 39050 | |
| }, | |
| { | |
| "epoch": 11.379890544946436, | |
| "grad_norm": 0.35445570945739746, | |
| "learning_rate": 0.0004637192778101339, | |
| "loss": 3.3369, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 11.394445738239403, | |
| "grad_norm": 0.3597460985183716, | |
| "learning_rate": 0.0004635445544554455, | |
| "loss": 3.3324, | |
| "step": 39150 | |
| }, | |
| { | |
| "epoch": 11.40900093153237, | |
| "grad_norm": 0.33990415930747986, | |
| "learning_rate": 0.0004633698311007571, | |
| "loss": 3.3461, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 11.423556124825337, | |
| "grad_norm": 0.3448801040649414, | |
| "learning_rate": 0.0004631951077460687, | |
| "loss": 3.3306, | |
| "step": 39250 | |
| }, | |
| { | |
| "epoch": 11.438111318118304, | |
| "grad_norm": 0.3563622534275055, | |
| "learning_rate": 0.00046302038439138024, | |
| "loss": 3.336, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 11.452666511411271, | |
| "grad_norm": 0.3600659668445587, | |
| "learning_rate": 0.0004628456610366919, | |
| "loss": 3.3442, | |
| "step": 39350 | |
| }, | |
| { | |
| "epoch": 11.467221704704238, | |
| "grad_norm": 0.32568398118019104, | |
| "learning_rate": 0.0004626709376820035, | |
| "loss": 3.3392, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 11.481776897997205, | |
| "grad_norm": 0.35172879695892334, | |
| "learning_rate": 0.0004624962143273151, | |
| "loss": 3.3516, | |
| "step": 39450 | |
| }, | |
| { | |
| "epoch": 11.496332091290173, | |
| "grad_norm": 0.3447670638561249, | |
| "learning_rate": 0.0004623214909726266, | |
| "loss": 3.3333, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 11.51088728458314, | |
| "grad_norm": 0.34315159916877747, | |
| "learning_rate": 0.0004621467676179382, | |
| "loss": 3.3565, | |
| "step": 39550 | |
| }, | |
| { | |
| "epoch": 11.525442477876107, | |
| "grad_norm": 0.3695205748081207, | |
| "learning_rate": 0.0004619720442632498, | |
| "loss": 3.3595, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 11.539997671169074, | |
| "grad_norm": 0.34323710203170776, | |
| "learning_rate": 0.00046179732090856145, | |
| "loss": 3.3469, | |
| "step": 39650 | |
| }, | |
| { | |
| "epoch": 11.55455286446204, | |
| "grad_norm": 0.33781781792640686, | |
| "learning_rate": 0.000461622597553873, | |
| "loss": 3.3469, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 11.569108057755006, | |
| "grad_norm": 0.3478993773460388, | |
| "learning_rate": 0.0004614478741991846, | |
| "loss": 3.3564, | |
| "step": 39750 | |
| }, | |
| { | |
| "epoch": 11.583663251047973, | |
| "grad_norm": 0.34204310178756714, | |
| "learning_rate": 0.0004612731508444962, | |
| "loss": 3.3526, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 11.59821844434094, | |
| "grad_norm": 0.36109477281570435, | |
| "learning_rate": 0.0004610984274898077, | |
| "loss": 3.3517, | |
| "step": 39850 | |
| }, | |
| { | |
| "epoch": 11.612773637633907, | |
| "grad_norm": 0.34194666147232056, | |
| "learning_rate": 0.00046092370413511937, | |
| "loss": 3.3634, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 11.627328830926874, | |
| "grad_norm": 0.3449431359767914, | |
| "learning_rate": 0.00046074898078043096, | |
| "loss": 3.3512, | |
| "step": 39950 | |
| }, | |
| { | |
| "epoch": 11.641884024219841, | |
| "grad_norm": 0.3200359046459198, | |
| "learning_rate": 0.00046057425742574256, | |
| "loss": 3.3651, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.641884024219841, | |
| "eval_accuracy": 0.3697300903162084, | |
| "eval_loss": 3.5560121536254883, | |
| "eval_runtime": 181.1724, | |
| "eval_samples_per_second": 91.907, | |
| "eval_steps_per_second": 5.746, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.656439217512808, | |
| "grad_norm": 0.3647145628929138, | |
| "learning_rate": 0.0004603995340710541, | |
| "loss": 3.3446, | |
| "step": 40050 | |
| }, | |
| { | |
| "epoch": 11.670994410805775, | |
| "grad_norm": 0.37295401096343994, | |
| "learning_rate": 0.0004602248107163657, | |
| "loss": 3.3714, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 11.685549604098743, | |
| "grad_norm": 0.3441055119037628, | |
| "learning_rate": 0.0004600500873616773, | |
| "loss": 3.3631, | |
| "step": 40150 | |
| }, | |
| { | |
| "epoch": 11.70010479739171, | |
| "grad_norm": 0.3755488097667694, | |
| "learning_rate": 0.00045987536400698894, | |
| "loss": 3.3652, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 11.714659990684677, | |
| "grad_norm": 0.3308577239513397, | |
| "learning_rate": 0.0004597006406523005, | |
| "loss": 3.367, | |
| "step": 40250 | |
| }, | |
| { | |
| "epoch": 11.729215183977644, | |
| "grad_norm": 0.3554903268814087, | |
| "learning_rate": 0.00045952591729761207, | |
| "loss": 3.3616, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 11.74377037727061, | |
| "grad_norm": 0.3407958745956421, | |
| "learning_rate": 0.00045935119394292367, | |
| "loss": 3.3621, | |
| "step": 40350 | |
| }, | |
| { | |
| "epoch": 11.758325570563578, | |
| "grad_norm": 0.3670002520084381, | |
| "learning_rate": 0.00045917647058823526, | |
| "loss": 3.3642, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 11.772880763856545, | |
| "grad_norm": 0.3373028337955475, | |
| "learning_rate": 0.0004590017472335468, | |
| "loss": 3.3577, | |
| "step": 40450 | |
| }, | |
| { | |
| "epoch": 11.78743595714951, | |
| "grad_norm": 0.3528793752193451, | |
| "learning_rate": 0.00045882702387885845, | |
| "loss": 3.3686, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 11.801991150442477, | |
| "grad_norm": 0.3317636549472809, | |
| "learning_rate": 0.00045865230052417004, | |
| "loss": 3.3664, | |
| "step": 40550 | |
| }, | |
| { | |
| "epoch": 11.816546343735444, | |
| "grad_norm": 0.3258965313434601, | |
| "learning_rate": 0.00045847757716948164, | |
| "loss": 3.3817, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 11.831101537028411, | |
| "grad_norm": 0.348578542470932, | |
| "learning_rate": 0.0004583028538147932, | |
| "loss": 3.3611, | |
| "step": 40650 | |
| }, | |
| { | |
| "epoch": 11.845656730321378, | |
| "grad_norm": 0.3560961186885834, | |
| "learning_rate": 0.00045812813046010477, | |
| "loss": 3.3698, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 11.860211923614346, | |
| "grad_norm": 0.3377649486064911, | |
| "learning_rate": 0.0004579534071054164, | |
| "loss": 3.3704, | |
| "step": 40750 | |
| }, | |
| { | |
| "epoch": 11.874767116907313, | |
| "grad_norm": 0.33449798822402954, | |
| "learning_rate": 0.000457778683750728, | |
| "loss": 3.3793, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 11.88932231020028, | |
| "grad_norm": 0.3655473589897156, | |
| "learning_rate": 0.00045760396039603955, | |
| "loss": 3.3721, | |
| "step": 40850 | |
| }, | |
| { | |
| "epoch": 11.903877503493247, | |
| "grad_norm": 0.3865300416946411, | |
| "learning_rate": 0.00045742923704135115, | |
| "loss": 3.3725, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 11.918432696786214, | |
| "grad_norm": 0.3576103150844574, | |
| "learning_rate": 0.00045725451368666274, | |
| "loss": 3.3717, | |
| "step": 40950 | |
| }, | |
| { | |
| "epoch": 11.93298789007918, | |
| "grad_norm": 0.32919180393218994, | |
| "learning_rate": 0.0004570797903319743, | |
| "loss": 3.3767, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.93298789007918, | |
| "eval_accuracy": 0.370469572121865, | |
| "eval_loss": 3.5477712154388428, | |
| "eval_runtime": 181.325, | |
| "eval_samples_per_second": 91.83, | |
| "eval_steps_per_second": 5.741, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.947543083372148, | |
| "grad_norm": 0.32170921564102173, | |
| "learning_rate": 0.00045690506697728593, | |
| "loss": 3.3728, | |
| "step": 41050 | |
| }, | |
| { | |
| "epoch": 11.962098276665115, | |
| "grad_norm": 0.3520164489746094, | |
| "learning_rate": 0.0004567303436225975, | |
| "loss": 3.37, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 11.97665346995808, | |
| "grad_norm": 0.34390193223953247, | |
| "learning_rate": 0.0004565556202679091, | |
| "loss": 3.3699, | |
| "step": 41150 | |
| }, | |
| { | |
| "epoch": 11.991208663251047, | |
| "grad_norm": 0.341876357793808, | |
| "learning_rate": 0.00045638089691322066, | |
| "loss": 3.3689, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 12.005530973451327, | |
| "grad_norm": 0.36311545968055725, | |
| "learning_rate": 0.00045620617355853225, | |
| "loss": 3.3312, | |
| "step": 41250 | |
| }, | |
| { | |
| "epoch": 12.020086166744294, | |
| "grad_norm": 0.3760519325733185, | |
| "learning_rate": 0.0004560314502038439, | |
| "loss": 3.2632, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 12.034641360037261, | |
| "grad_norm": 0.35626885294914246, | |
| "learning_rate": 0.0004558567268491555, | |
| "loss": 3.2631, | |
| "step": 41350 | |
| }, | |
| { | |
| "epoch": 12.049196553330228, | |
| "grad_norm": 0.3312586545944214, | |
| "learning_rate": 0.00045568200349446704, | |
| "loss": 3.2653, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 12.063751746623195, | |
| "grad_norm": 0.33731192350387573, | |
| "learning_rate": 0.00045550728013977863, | |
| "loss": 3.2639, | |
| "step": 41450 | |
| }, | |
| { | |
| "epoch": 12.078306939916162, | |
| "grad_norm": 0.3921743929386139, | |
| "learning_rate": 0.0004553325567850902, | |
| "loss": 3.2823, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 12.09286213320913, | |
| "grad_norm": 0.35110893845558167, | |
| "learning_rate": 0.0004551578334304018, | |
| "loss": 3.2822, | |
| "step": 41550 | |
| }, | |
| { | |
| "epoch": 12.107417326502096, | |
| "grad_norm": 0.32425668835639954, | |
| "learning_rate": 0.00045498311007571347, | |
| "loss": 3.2891, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 12.121972519795063, | |
| "grad_norm": 0.35301515460014343, | |
| "learning_rate": 0.000454808386721025, | |
| "loss": 3.2916, | |
| "step": 41650 | |
| }, | |
| { | |
| "epoch": 12.13652771308803, | |
| "grad_norm": 0.35211730003356934, | |
| "learning_rate": 0.0004546336633663366, | |
| "loss": 3.2885, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 12.151082906380998, | |
| "grad_norm": 0.35022473335266113, | |
| "learning_rate": 0.0004544589400116482, | |
| "loss": 3.2871, | |
| "step": 41750 | |
| }, | |
| { | |
| "epoch": 12.165638099673963, | |
| "grad_norm": 0.33055204153060913, | |
| "learning_rate": 0.00045428421665695974, | |
| "loss": 3.2922, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 12.18019329296693, | |
| "grad_norm": 0.3559176027774811, | |
| "learning_rate": 0.00045410949330227133, | |
| "loss": 3.301, | |
| "step": 41850 | |
| }, | |
| { | |
| "epoch": 12.194748486259897, | |
| "grad_norm": 0.3754260838031769, | |
| "learning_rate": 0.000453934769947583, | |
| "loss": 3.3022, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 12.209303679552864, | |
| "grad_norm": 0.33595168590545654, | |
| "learning_rate": 0.0004537600465928946, | |
| "loss": 3.3005, | |
| "step": 41950 | |
| }, | |
| { | |
| "epoch": 12.223858872845831, | |
| "grad_norm": 0.3507530987262726, | |
| "learning_rate": 0.0004535853232382061, | |
| "loss": 3.3124, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.223858872845831, | |
| "eval_accuracy": 0.3698600755159948, | |
| "eval_loss": 3.563331127166748, | |
| "eval_runtime": 181.337, | |
| "eval_samples_per_second": 91.824, | |
| "eval_steps_per_second": 5.741, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.238414066138798, | |
| "grad_norm": 0.35963207483291626, | |
| "learning_rate": 0.0004534105998835177, | |
| "loss": 3.3003, | |
| "step": 42050 | |
| }, | |
| { | |
| "epoch": 12.252969259431765, | |
| "grad_norm": 0.3779832124710083, | |
| "learning_rate": 0.0004532358765288293, | |
| "loss": 3.297, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 12.267524452724732, | |
| "grad_norm": 0.38168567419052124, | |
| "learning_rate": 0.00045306115317414095, | |
| "loss": 3.3092, | |
| "step": 42150 | |
| }, | |
| { | |
| "epoch": 12.2820796460177, | |
| "grad_norm": 0.3286755681037903, | |
| "learning_rate": 0.0004528864298194525, | |
| "loss": 3.311, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 12.296634839310666, | |
| "grad_norm": 0.3363521993160248, | |
| "learning_rate": 0.0004527117064647641, | |
| "loss": 3.3073, | |
| "step": 42250 | |
| }, | |
| { | |
| "epoch": 12.311190032603633, | |
| "grad_norm": 0.35366666316986084, | |
| "learning_rate": 0.0004525369831100757, | |
| "loss": 3.303, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 12.3257452258966, | |
| "grad_norm": 0.35679879784584045, | |
| "learning_rate": 0.0004523622597553872, | |
| "loss": 3.331, | |
| "step": 42350 | |
| }, | |
| { | |
| "epoch": 12.340300419189568, | |
| "grad_norm": 0.35085782408714294, | |
| "learning_rate": 0.0004521875364006988, | |
| "loss": 3.31, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 12.354855612482535, | |
| "grad_norm": 0.36564141511917114, | |
| "learning_rate": 0.00045201281304601046, | |
| "loss": 3.3175, | |
| "step": 42450 | |
| }, | |
| { | |
| "epoch": 12.3694108057755, | |
| "grad_norm": 0.3681644797325134, | |
| "learning_rate": 0.00045183808969132206, | |
| "loss": 3.3174, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 12.383965999068467, | |
| "grad_norm": 0.363091379404068, | |
| "learning_rate": 0.00045166336633663365, | |
| "loss": 3.3259, | |
| "step": 42550 | |
| }, | |
| { | |
| "epoch": 12.398521192361434, | |
| "grad_norm": 0.38171902298927307, | |
| "learning_rate": 0.0004514886429819452, | |
| "loss": 3.314, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 12.413076385654401, | |
| "grad_norm": 0.3879528343677521, | |
| "learning_rate": 0.0004513139196272568, | |
| "loss": 3.3278, | |
| "step": 42650 | |
| }, | |
| { | |
| "epoch": 12.427631578947368, | |
| "grad_norm": 0.34305742383003235, | |
| "learning_rate": 0.00045113919627256843, | |
| "loss": 3.3164, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 12.442186772240335, | |
| "grad_norm": 0.3577149510383606, | |
| "learning_rate": 0.00045096447291788003, | |
| "loss": 3.3243, | |
| "step": 42750 | |
| }, | |
| { | |
| "epoch": 12.456741965533302, | |
| "grad_norm": 0.33020853996276855, | |
| "learning_rate": 0.00045078974956319157, | |
| "loss": 3.3335, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 12.47129715882627, | |
| "grad_norm": 0.36573347449302673, | |
| "learning_rate": 0.00045061502620850316, | |
| "loss": 3.3238, | |
| "step": 42850 | |
| }, | |
| { | |
| "epoch": 12.485852352119236, | |
| "grad_norm": 0.3603604733943939, | |
| "learning_rate": 0.00045044030285381476, | |
| "loss": 3.3252, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 12.500407545412203, | |
| "grad_norm": 0.3663039207458496, | |
| "learning_rate": 0.0004502655794991263, | |
| "loss": 3.3292, | |
| "step": 42950 | |
| }, | |
| { | |
| "epoch": 12.51496273870517, | |
| "grad_norm": 0.33189836144447327, | |
| "learning_rate": 0.00045009085614443795, | |
| "loss": 3.3323, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.51496273870517, | |
| "eval_accuracy": 0.37042279625431074, | |
| "eval_loss": 3.5531504154205322, | |
| "eval_runtime": 181.3654, | |
| "eval_samples_per_second": 91.809, | |
| "eval_steps_per_second": 5.74, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.529517931998138, | |
| "grad_norm": 0.39830282330513, | |
| "learning_rate": 0.00044991613278974954, | |
| "loss": 3.3356, | |
| "step": 43050 | |
| }, | |
| { | |
| "epoch": 12.544073125291105, | |
| "grad_norm": 0.3498653173446655, | |
| "learning_rate": 0.00044974140943506113, | |
| "loss": 3.3376, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 12.55862831858407, | |
| "grad_norm": 0.3489161431789398, | |
| "learning_rate": 0.0004495666860803727, | |
| "loss": 3.3268, | |
| "step": 43150 | |
| }, | |
| { | |
| "epoch": 12.573183511877037, | |
| "grad_norm": 0.36358317732810974, | |
| "learning_rate": 0.00044939196272568427, | |
| "loss": 3.334, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 12.587738705170004, | |
| "grad_norm": 0.35750314593315125, | |
| "learning_rate": 0.00044921723937099586, | |
| "loss": 3.332, | |
| "step": 43250 | |
| }, | |
| { | |
| "epoch": 12.602293898462971, | |
| "grad_norm": 0.3437962830066681, | |
| "learning_rate": 0.0004490425160163075, | |
| "loss": 3.3463, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 12.616849091755938, | |
| "grad_norm": 0.36788836121559143, | |
| "learning_rate": 0.00044886779266161905, | |
| "loss": 3.3322, | |
| "step": 43350 | |
| }, | |
| { | |
| "epoch": 12.631404285048905, | |
| "grad_norm": 0.3601713180541992, | |
| "learning_rate": 0.00044869306930693065, | |
| "loss": 3.3432, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 12.645959478341872, | |
| "grad_norm": 0.3093632459640503, | |
| "learning_rate": 0.00044851834595224224, | |
| "loss": 3.3397, | |
| "step": 43450 | |
| }, | |
| { | |
| "epoch": 12.66051467163484, | |
| "grad_norm": 0.34150272607803345, | |
| "learning_rate": 0.00044834362259755383, | |
| "loss": 3.3498, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 12.675069864927806, | |
| "grad_norm": 0.3343871533870697, | |
| "learning_rate": 0.00044816889924286543, | |
| "loss": 3.3388, | |
| "step": 43550 | |
| }, | |
| { | |
| "epoch": 12.689625058220773, | |
| "grad_norm": 0.33877092599868774, | |
| "learning_rate": 0.000447994175888177, | |
| "loss": 3.3481, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 12.70418025151374, | |
| "grad_norm": 0.35952886939048767, | |
| "learning_rate": 0.0004478194525334886, | |
| "loss": 3.3494, | |
| "step": 43650 | |
| }, | |
| { | |
| "epoch": 12.718735444806708, | |
| "grad_norm": 0.3660774827003479, | |
| "learning_rate": 0.0004476447291788002, | |
| "loss": 3.3384, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 12.733290638099675, | |
| "grad_norm": 0.35078608989715576, | |
| "learning_rate": 0.00044747000582411175, | |
| "loss": 3.3374, | |
| "step": 43750 | |
| }, | |
| { | |
| "epoch": 12.747845831392642, | |
| "grad_norm": 0.3302287459373474, | |
| "learning_rate": 0.00044729528246942335, | |
| "loss": 3.3384, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 12.762401024685607, | |
| "grad_norm": 0.3574998378753662, | |
| "learning_rate": 0.000447120559114735, | |
| "loss": 3.3324, | |
| "step": 43850 | |
| }, | |
| { | |
| "epoch": 12.776956217978574, | |
| "grad_norm": 0.3275946080684662, | |
| "learning_rate": 0.0004469458357600466, | |
| "loss": 3.3458, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 12.791511411271541, | |
| "grad_norm": 0.3478822112083435, | |
| "learning_rate": 0.00044677111240535813, | |
| "loss": 3.3481, | |
| "step": 43950 | |
| }, | |
| { | |
| "epoch": 12.806066604564508, | |
| "grad_norm": 0.3444738984107971, | |
| "learning_rate": 0.0004465963890506697, | |
| "loss": 3.3498, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.806066604564508, | |
| "eval_accuracy": 0.37071473408095584, | |
| "eval_loss": 3.550508975982666, | |
| "eval_runtime": 181.4057, | |
| "eval_samples_per_second": 91.789, | |
| "eval_steps_per_second": 5.739, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.820621797857475, | |
| "grad_norm": 0.3442026674747467, | |
| "learning_rate": 0.0004464216656959813, | |
| "loss": 3.3616, | |
| "step": 44050 | |
| }, | |
| { | |
| "epoch": 12.835176991150442, | |
| "grad_norm": 0.3639858067035675, | |
| "learning_rate": 0.00044624694234129297, | |
| "loss": 3.3583, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 12.84973218444341, | |
| "grad_norm": 0.35911089181900024, | |
| "learning_rate": 0.0004460722189866045, | |
| "loss": 3.3487, | |
| "step": 44150 | |
| }, | |
| { | |
| "epoch": 12.864287377736376, | |
| "grad_norm": 0.3291614353656769, | |
| "learning_rate": 0.0004458974956319161, | |
| "loss": 3.3515, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 12.878842571029343, | |
| "grad_norm": 0.3494235575199127, | |
| "learning_rate": 0.0004457227722772277, | |
| "loss": 3.3511, | |
| "step": 44250 | |
| }, | |
| { | |
| "epoch": 12.89339776432231, | |
| "grad_norm": 0.3374604880809784, | |
| "learning_rate": 0.00044554804892253923, | |
| "loss": 3.35, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 12.907952957615278, | |
| "grad_norm": 0.3468201160430908, | |
| "learning_rate": 0.00044537332556785083, | |
| "loss": 3.3465, | |
| "step": 44350 | |
| }, | |
| { | |
| "epoch": 12.922508150908245, | |
| "grad_norm": 0.35529449582099915, | |
| "learning_rate": 0.0004451986022131625, | |
| "loss": 3.3596, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 12.937063344201212, | |
| "grad_norm": 0.3489094376564026, | |
| "learning_rate": 0.00044502387885847407, | |
| "loss": 3.3567, | |
| "step": 44450 | |
| }, | |
| { | |
| "epoch": 12.951618537494177, | |
| "grad_norm": 0.37239301204681396, | |
| "learning_rate": 0.0004448491555037856, | |
| "loss": 3.3536, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 12.966173730787144, | |
| "grad_norm": 0.40329527854919434, | |
| "learning_rate": 0.0004446744321490972, | |
| "loss": 3.3503, | |
| "step": 44550 | |
| }, | |
| { | |
| "epoch": 12.980728924080111, | |
| "grad_norm": 0.3665720820426941, | |
| "learning_rate": 0.0004444997087944088, | |
| "loss": 3.3592, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 12.995284117373078, | |
| "grad_norm": 0.3658985197544098, | |
| "learning_rate": 0.0004443249854397204, | |
| "loss": 3.3493, | |
| "step": 44650 | |
| }, | |
| { | |
| "epoch": 13.009606427573358, | |
| "grad_norm": 0.34227049350738525, | |
| "learning_rate": 0.000444150262085032, | |
| "loss": 3.2852, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 13.024161620866325, | |
| "grad_norm": 0.3462371230125427, | |
| "learning_rate": 0.0004439755387303436, | |
| "loss": 3.2481, | |
| "step": 44750 | |
| }, | |
| { | |
| "epoch": 13.038716814159292, | |
| "grad_norm": 0.36165347695350647, | |
| "learning_rate": 0.0004438008153756552, | |
| "loss": 3.2525, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 13.053272007452259, | |
| "grad_norm": 0.3508550822734833, | |
| "learning_rate": 0.00044362609202096677, | |
| "loss": 3.259, | |
| "step": 44850 | |
| }, | |
| { | |
| "epoch": 13.067827200745226, | |
| "grad_norm": 0.3499922752380371, | |
| "learning_rate": 0.0004434513686662783, | |
| "loss": 3.2469, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 13.082382394038193, | |
| "grad_norm": 0.3820374011993408, | |
| "learning_rate": 0.00044327664531158996, | |
| "loss": 3.2486, | |
| "step": 44950 | |
| }, | |
| { | |
| "epoch": 13.09693758733116, | |
| "grad_norm": 0.34704211354255676, | |
| "learning_rate": 0.00044310192195690155, | |
| "loss": 3.2658, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.09693758733116, | |
| "eval_accuracy": 0.37041080846915864, | |
| "eval_loss": 3.5606274604797363, | |
| "eval_runtime": 181.4519, | |
| "eval_samples_per_second": 91.765, | |
| "eval_steps_per_second": 5.737, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.111492780624127, | |
| "grad_norm": 0.3327675461769104, | |
| "learning_rate": 0.00044292719860221315, | |
| "loss": 3.265, | |
| "step": 45050 | |
| }, | |
| { | |
| "epoch": 13.126047973917094, | |
| "grad_norm": 0.32860827445983887, | |
| "learning_rate": 0.0004427524752475247, | |
| "loss": 3.2626, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 13.140603167210061, | |
| "grad_norm": 0.3338595926761627, | |
| "learning_rate": 0.0004425777518928363, | |
| "loss": 3.2655, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 13.155158360503027, | |
| "grad_norm": 0.3572806715965271, | |
| "learning_rate": 0.0004424030285381479, | |
| "loss": 3.2779, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 13.169713553795994, | |
| "grad_norm": 0.36495622992515564, | |
| "learning_rate": 0.0004422283051834595, | |
| "loss": 3.2692, | |
| "step": 45250 | |
| }, | |
| { | |
| "epoch": 13.18426874708896, | |
| "grad_norm": 0.39937713742256165, | |
| "learning_rate": 0.00044205358182877107, | |
| "loss": 3.2933, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 13.198823940381928, | |
| "grad_norm": 0.3623490333557129, | |
| "learning_rate": 0.00044187885847408266, | |
| "loss": 3.2924, | |
| "step": 45350 | |
| }, | |
| { | |
| "epoch": 13.213379133674895, | |
| "grad_norm": 0.3873792588710785, | |
| "learning_rate": 0.00044170413511939425, | |
| "loss": 3.2905, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 13.227934326967862, | |
| "grad_norm": 0.36040782928466797, | |
| "learning_rate": 0.0004415294117647058, | |
| "loss": 3.2894, | |
| "step": 45450 | |
| }, | |
| { | |
| "epoch": 13.242489520260829, | |
| "grad_norm": 0.3621423542499542, | |
| "learning_rate": 0.00044135468841001744, | |
| "loss": 3.2964, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 13.257044713553796, | |
| "grad_norm": 0.3496916592121124, | |
| "learning_rate": 0.00044117996505532904, | |
| "loss": 3.2837, | |
| "step": 45550 | |
| }, | |
| { | |
| "epoch": 13.271599906846763, | |
| "grad_norm": 0.4022481441497803, | |
| "learning_rate": 0.00044100524170064063, | |
| "loss": 3.2878, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 13.28615510013973, | |
| "grad_norm": 0.38104742765426636, | |
| "learning_rate": 0.0004408305183459522, | |
| "loss": 3.2934, | |
| "step": 45650 | |
| }, | |
| { | |
| "epoch": 13.300710293432697, | |
| "grad_norm": 0.37204158306121826, | |
| "learning_rate": 0.00044065579499126377, | |
| "loss": 3.2973, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 13.315265486725664, | |
| "grad_norm": 0.3469567894935608, | |
| "learning_rate": 0.00044048107163657536, | |
| "loss": 3.305, | |
| "step": 45750 | |
| }, | |
| { | |
| "epoch": 13.329820680018631, | |
| "grad_norm": 0.34882915019989014, | |
| "learning_rate": 0.000440306348281887, | |
| "loss": 3.3061, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 13.344375873311598, | |
| "grad_norm": 0.37024763226509094, | |
| "learning_rate": 0.0004401316249271986, | |
| "loss": 3.3038, | |
| "step": 45850 | |
| }, | |
| { | |
| "epoch": 13.358931066604564, | |
| "grad_norm": 0.37024813890457153, | |
| "learning_rate": 0.00043995690157251014, | |
| "loss": 3.3044, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 13.37348625989753, | |
| "grad_norm": 0.36839035153388977, | |
| "learning_rate": 0.00043978217821782174, | |
| "loss": 3.3079, | |
| "step": 45950 | |
| }, | |
| { | |
| "epoch": 13.388041453190498, | |
| "grad_norm": 0.3722715377807617, | |
| "learning_rate": 0.00043960745486313333, | |
| "loss": 3.3054, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.388041453190498, | |
| "eval_accuracy": 0.3708863239468584, | |
| "eval_loss": 3.5549628734588623, | |
| "eval_runtime": 181.4286, | |
| "eval_samples_per_second": 91.777, | |
| "eval_steps_per_second": 5.738, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.402596646483465, | |
| "grad_norm": 0.3347816467285156, | |
| "learning_rate": 0.00043943273150844487, | |
| "loss": 3.3029, | |
| "step": 46050 | |
| }, | |
| { | |
| "epoch": 13.417151839776432, | |
| "grad_norm": 0.3504791259765625, | |
| "learning_rate": 0.0004392580081537565, | |
| "loss": 3.3015, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 13.431707033069399, | |
| "grad_norm": 0.37427136301994324, | |
| "learning_rate": 0.0004390832847990681, | |
| "loss": 3.3096, | |
| "step": 46150 | |
| }, | |
| { | |
| "epoch": 13.446262226362366, | |
| "grad_norm": 0.3400617241859436, | |
| "learning_rate": 0.0004389085614443797, | |
| "loss": 3.307, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 13.460817419655333, | |
| "grad_norm": 0.37000834941864014, | |
| "learning_rate": 0.00043873383808969125, | |
| "loss": 3.2937, | |
| "step": 46250 | |
| }, | |
| { | |
| "epoch": 13.4753726129483, | |
| "grad_norm": 0.34978291392326355, | |
| "learning_rate": 0.00043855911473500284, | |
| "loss": 3.3096, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 13.489927806241267, | |
| "grad_norm": 0.3517734706401825, | |
| "learning_rate": 0.0004383843913803145, | |
| "loss": 3.3093, | |
| "step": 46350 | |
| }, | |
| { | |
| "epoch": 13.504482999534234, | |
| "grad_norm": 0.3470326364040375, | |
| "learning_rate": 0.0004382096680256261, | |
| "loss": 3.3038, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 13.519038192827201, | |
| "grad_norm": 0.37263551354408264, | |
| "learning_rate": 0.0004380349446709376, | |
| "loss": 3.3167, | |
| "step": 46450 | |
| }, | |
| { | |
| "epoch": 13.533593386120168, | |
| "grad_norm": 0.3580203056335449, | |
| "learning_rate": 0.0004378602213162492, | |
| "loss": 3.3184, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 13.548148579413134, | |
| "grad_norm": 0.3770522177219391, | |
| "learning_rate": 0.0004376854979615608, | |
| "loss": 3.3158, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 13.5627037727061, | |
| "grad_norm": 0.37715944647789, | |
| "learning_rate": 0.0004375107746068724, | |
| "loss": 3.3239, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 13.577258965999068, | |
| "grad_norm": 0.3532266616821289, | |
| "learning_rate": 0.000437336051252184, | |
| "loss": 3.3219, | |
| "step": 46650 | |
| }, | |
| { | |
| "epoch": 13.591814159292035, | |
| "grad_norm": 0.37203213572502136, | |
| "learning_rate": 0.0004371613278974956, | |
| "loss": 3.3188, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 13.606369352585002, | |
| "grad_norm": 0.370649129152298, | |
| "learning_rate": 0.0004369866045428072, | |
| "loss": 3.3148, | |
| "step": 46750 | |
| }, | |
| { | |
| "epoch": 13.620924545877969, | |
| "grad_norm": 0.3378678858280182, | |
| "learning_rate": 0.0004368118811881188, | |
| "loss": 3.3152, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 13.635479739170936, | |
| "grad_norm": 0.3273431360721588, | |
| "learning_rate": 0.0004366371578334303, | |
| "loss": 3.3226, | |
| "step": 46850 | |
| }, | |
| { | |
| "epoch": 13.650034932463903, | |
| "grad_norm": 0.349479079246521, | |
| "learning_rate": 0.000436462434478742, | |
| "loss": 3.3402, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 13.66459012575687, | |
| "grad_norm": 0.4112270772457123, | |
| "learning_rate": 0.00043628771112405357, | |
| "loss": 3.331, | |
| "step": 46950 | |
| }, | |
| { | |
| "epoch": 13.679145319049837, | |
| "grad_norm": 0.3594585061073303, | |
| "learning_rate": 0.00043611298776936516, | |
| "loss": 3.3295, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.679145319049837, | |
| "eval_accuracy": 0.3710210102388613, | |
| "eval_loss": 3.5495879650115967, | |
| "eval_runtime": 181.3002, | |
| "eval_samples_per_second": 91.842, | |
| "eval_steps_per_second": 5.742, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.693700512342804, | |
| "grad_norm": 0.3493228852748871, | |
| "learning_rate": 0.0004359382644146767, | |
| "loss": 3.3297, | |
| "step": 47050 | |
| }, | |
| { | |
| "epoch": 13.708255705635771, | |
| "grad_norm": 0.34904351830482483, | |
| "learning_rate": 0.0004357635410599883, | |
| "loss": 3.3279, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 13.722810898928739, | |
| "grad_norm": 0.34903162717819214, | |
| "learning_rate": 0.0004355888177052999, | |
| "loss": 3.3357, | |
| "step": 47150 | |
| }, | |
| { | |
| "epoch": 13.737366092221706, | |
| "grad_norm": 0.34454700350761414, | |
| "learning_rate": 0.00043541409435061154, | |
| "loss": 3.3259, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 13.75192128551467, | |
| "grad_norm": 0.35162457823753357, | |
| "learning_rate": 0.0004352393709959231, | |
| "loss": 3.3325, | |
| "step": 47250 | |
| }, | |
| { | |
| "epoch": 13.766476478807638, | |
| "grad_norm": 0.3408013880252838, | |
| "learning_rate": 0.0004350646476412347, | |
| "loss": 3.3389, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 13.781031672100605, | |
| "grad_norm": 0.354308545589447, | |
| "learning_rate": 0.00043488992428654627, | |
| "loss": 3.3408, | |
| "step": 47350 | |
| }, | |
| { | |
| "epoch": 13.795586865393572, | |
| "grad_norm": 0.40119293332099915, | |
| "learning_rate": 0.0004347152009318578, | |
| "loss": 3.3234, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 13.810142058686539, | |
| "grad_norm": 0.3471970558166504, | |
| "learning_rate": 0.00043454047757716946, | |
| "loss": 3.333, | |
| "step": 47450 | |
| }, | |
| { | |
| "epoch": 13.824697251979506, | |
| "grad_norm": 0.36737313866615295, | |
| "learning_rate": 0.00043436575422248105, | |
| "loss": 3.333, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 13.839252445272473, | |
| "grad_norm": 0.3637086749076843, | |
| "learning_rate": 0.00043419103086779265, | |
| "loss": 3.3328, | |
| "step": 47550 | |
| }, | |
| { | |
| "epoch": 13.85380763856544, | |
| "grad_norm": 0.3733203113079071, | |
| "learning_rate": 0.0004340163075131042, | |
| "loss": 3.3223, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 13.868362831858407, | |
| "grad_norm": 0.3271060883998871, | |
| "learning_rate": 0.0004338415841584158, | |
| "loss": 3.3405, | |
| "step": 47650 | |
| }, | |
| { | |
| "epoch": 13.882918025151374, | |
| "grad_norm": 0.35861730575561523, | |
| "learning_rate": 0.0004336668608037274, | |
| "loss": 3.3473, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 13.897473218444341, | |
| "grad_norm": 0.3527565598487854, | |
| "learning_rate": 0.000433492137449039, | |
| "loss": 3.3341, | |
| "step": 47750 | |
| }, | |
| { | |
| "epoch": 13.912028411737309, | |
| "grad_norm": 0.40213543176651, | |
| "learning_rate": 0.00043331741409435056, | |
| "loss": 3.3284, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 13.926583605030276, | |
| "grad_norm": 0.34782031178474426, | |
| "learning_rate": 0.00043314269073966216, | |
| "loss": 3.3332, | |
| "step": 47850 | |
| }, | |
| { | |
| "epoch": 13.94113879832324, | |
| "grad_norm": 0.3511146903038025, | |
| "learning_rate": 0.00043296796738497375, | |
| "loss": 3.3264, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 13.955693991616208, | |
| "grad_norm": 0.37038230895996094, | |
| "learning_rate": 0.00043279324403028535, | |
| "loss": 3.331, | |
| "step": 47950 | |
| }, | |
| { | |
| "epoch": 13.970249184909175, | |
| "grad_norm": 0.37031859159469604, | |
| "learning_rate": 0.0004326185206755969, | |
| "loss": 3.3345, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.970249184909175, | |
| "eval_accuracy": 0.37165318961467614, | |
| "eval_loss": 3.5406837463378906, | |
| "eval_runtime": 181.3711, | |
| "eval_samples_per_second": 91.806, | |
| "eval_steps_per_second": 5.74, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.984804378202142, | |
| "grad_norm": 0.3739323019981384, | |
| "learning_rate": 0.00043244379732090854, | |
| "loss": 3.3379, | |
| "step": 48050 | |
| }, | |
| { | |
| "epoch": 13.99935957149511, | |
| "grad_norm": 0.341511070728302, | |
| "learning_rate": 0.00043226907396622013, | |
| "loss": 3.3512, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 14.013681881695389, | |
| "grad_norm": 0.37125998735427856, | |
| "learning_rate": 0.0004320943506115317, | |
| "loss": 3.2204, | |
| "step": 48150 | |
| }, | |
| { | |
| "epoch": 14.028237074988356, | |
| "grad_norm": 0.3771876394748688, | |
| "learning_rate": 0.00043191962725684326, | |
| "loss": 3.2392, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 14.042792268281323, | |
| "grad_norm": 0.35821080207824707, | |
| "learning_rate": 0.00043174490390215486, | |
| "loss": 3.2436, | |
| "step": 48250 | |
| }, | |
| { | |
| "epoch": 14.05734746157429, | |
| "grad_norm": 0.370997816324234, | |
| "learning_rate": 0.0004315701805474665, | |
| "loss": 3.2457, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 14.071902654867257, | |
| "grad_norm": 0.3405769169330597, | |
| "learning_rate": 0.0004313954571927781, | |
| "loss": 3.2555, | |
| "step": 48350 | |
| }, | |
| { | |
| "epoch": 14.086457848160224, | |
| "grad_norm": 0.35191479325294495, | |
| "learning_rate": 0.00043122073383808964, | |
| "loss": 3.2478, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 14.101013041453191, | |
| "grad_norm": 0.34124019742012024, | |
| "learning_rate": 0.00043104601048340124, | |
| "loss": 3.2366, | |
| "step": 48450 | |
| }, | |
| { | |
| "epoch": 14.115568234746158, | |
| "grad_norm": 0.3625926375389099, | |
| "learning_rate": 0.00043087128712871283, | |
| "loss": 3.252, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 14.130123428039123, | |
| "grad_norm": 0.364780068397522, | |
| "learning_rate": 0.00043069656377402437, | |
| "loss": 3.2408, | |
| "step": 48550 | |
| }, | |
| { | |
| "epoch": 14.14467862133209, | |
| "grad_norm": 0.35787588357925415, | |
| "learning_rate": 0.000430521840419336, | |
| "loss": 3.255, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 14.159233814625058, | |
| "grad_norm": 0.3852957785129547, | |
| "learning_rate": 0.0004303471170646476, | |
| "loss": 3.2672, | |
| "step": 48650 | |
| }, | |
| { | |
| "epoch": 14.173789007918025, | |
| "grad_norm": 0.3464197516441345, | |
| "learning_rate": 0.0004301723937099592, | |
| "loss": 3.2624, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 14.188344201210992, | |
| "grad_norm": 0.36321866512298584, | |
| "learning_rate": 0.00042999767035527075, | |
| "loss": 3.2599, | |
| "step": 48750 | |
| }, | |
| { | |
| "epoch": 14.202899394503959, | |
| "grad_norm": 0.3656206429004669, | |
| "learning_rate": 0.00042982294700058234, | |
| "loss": 3.2667, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 14.217454587796926, | |
| "grad_norm": 0.3317374885082245, | |
| "learning_rate": 0.000429648223645894, | |
| "loss": 3.2667, | |
| "step": 48850 | |
| }, | |
| { | |
| "epoch": 14.232009781089893, | |
| "grad_norm": 0.34478551149368286, | |
| "learning_rate": 0.0004294735002912056, | |
| "loss": 3.2794, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 14.24656497438286, | |
| "grad_norm": 0.38702359795570374, | |
| "learning_rate": 0.0004292987769365172, | |
| "loss": 3.2622, | |
| "step": 48950 | |
| }, | |
| { | |
| "epoch": 14.261120167675827, | |
| "grad_norm": 0.4132826030254364, | |
| "learning_rate": 0.0004291240535818287, | |
| "loss": 3.2725, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.261120167675827, | |
| "eval_accuracy": 0.37083684495127966, | |
| "eval_loss": 3.559509515762329, | |
| "eval_runtime": 181.4988, | |
| "eval_samples_per_second": 91.742, | |
| "eval_steps_per_second": 5.736, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.275675360968794, | |
| "grad_norm": 0.3839230537414551, | |
| "learning_rate": 0.0004289493302271403, | |
| "loss": 3.2781, | |
| "step": 49050 | |
| }, | |
| { | |
| "epoch": 14.290230554261761, | |
| "grad_norm": 0.3690083920955658, | |
| "learning_rate": 0.0004287746068724519, | |
| "loss": 3.2722, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 14.304785747554728, | |
| "grad_norm": 0.3392127752304077, | |
| "learning_rate": 0.00042859988351776356, | |
| "loss": 3.2868, | |
| "step": 49150 | |
| }, | |
| { | |
| "epoch": 14.319340940847695, | |
| "grad_norm": 0.3598864674568176, | |
| "learning_rate": 0.0004284251601630751, | |
| "loss": 3.2953, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 14.33389613414066, | |
| "grad_norm": 0.3549182116985321, | |
| "learning_rate": 0.0004282504368083867, | |
| "loss": 3.2795, | |
| "step": 49250 | |
| }, | |
| { | |
| "epoch": 14.348451327433628, | |
| "grad_norm": 0.3696673512458801, | |
| "learning_rate": 0.0004280757134536983, | |
| "loss": 3.2852, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 14.363006520726595, | |
| "grad_norm": 0.3530607521533966, | |
| "learning_rate": 0.0004279009900990098, | |
| "loss": 3.288, | |
| "step": 49350 | |
| }, | |
| { | |
| "epoch": 14.377561714019562, | |
| "grad_norm": 0.37303102016448975, | |
| "learning_rate": 0.0004277262667443214, | |
| "loss": 3.2851, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 14.392116907312529, | |
| "grad_norm": 0.3705254793167114, | |
| "learning_rate": 0.00042755154338963307, | |
| "loss": 3.2887, | |
| "step": 49450 | |
| }, | |
| { | |
| "epoch": 14.406672100605496, | |
| "grad_norm": 0.3585062623023987, | |
| "learning_rate": 0.00042737682003494466, | |
| "loss": 3.293, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 14.421227293898463, | |
| "grad_norm": 0.3528754711151123, | |
| "learning_rate": 0.0004272020966802562, | |
| "loss": 3.2798, | |
| "step": 49550 | |
| }, | |
| { | |
| "epoch": 14.43578248719143, | |
| "grad_norm": 0.3506697714328766, | |
| "learning_rate": 0.0004270273733255678, | |
| "loss": 3.2775, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 14.450337680484397, | |
| "grad_norm": 0.3489326238632202, | |
| "learning_rate": 0.0004268526499708794, | |
| "loss": 3.2826, | |
| "step": 49650 | |
| }, | |
| { | |
| "epoch": 14.464892873777364, | |
| "grad_norm": 0.3633829653263092, | |
| "learning_rate": 0.00042667792661619104, | |
| "loss": 3.2986, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 14.479448067070331, | |
| "grad_norm": 0.38335442543029785, | |
| "learning_rate": 0.0004265032032615026, | |
| "loss": 3.2962, | |
| "step": 49750 | |
| }, | |
| { | |
| "epoch": 14.494003260363298, | |
| "grad_norm": 0.3436025083065033, | |
| "learning_rate": 0.0004263284799068142, | |
| "loss": 3.2912, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 14.508558453656265, | |
| "grad_norm": 0.3850632905960083, | |
| "learning_rate": 0.00042615375655212577, | |
| "loss": 3.3004, | |
| "step": 49850 | |
| }, | |
| { | |
| "epoch": 14.52311364694923, | |
| "grad_norm": 0.34339576959609985, | |
| "learning_rate": 0.00042597903319743736, | |
| "loss": 3.2974, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 14.537668840242198, | |
| "grad_norm": 0.3751101791858673, | |
| "learning_rate": 0.0004258043098427489, | |
| "loss": 3.3055, | |
| "step": 49950 | |
| }, | |
| { | |
| "epoch": 14.552224033535165, | |
| "grad_norm": 0.4216310679912567, | |
| "learning_rate": 0.00042562958648806055, | |
| "loss": 3.3044, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.552224033535165, | |
| "eval_accuracy": 0.37140920292863944, | |
| "eval_loss": 3.5514838695526123, | |
| "eval_runtime": 181.3812, | |
| "eval_samples_per_second": 91.801, | |
| "eval_steps_per_second": 5.739, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.566779226828132, | |
| "grad_norm": 0.34719032049179077, | |
| "learning_rate": 0.00042545486313337214, | |
| "loss": 3.307, | |
| "step": 50050 | |
| }, | |
| { | |
| "epoch": 14.581334420121099, | |
| "grad_norm": 0.3421182930469513, | |
| "learning_rate": 0.00042528013977868374, | |
| "loss": 3.2997, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 14.595889613414066, | |
| "grad_norm": 0.3395073711872101, | |
| "learning_rate": 0.0004251054164239953, | |
| "loss": 3.3123, | |
| "step": 50150 | |
| }, | |
| { | |
| "epoch": 14.610444806707033, | |
| "grad_norm": 0.39163294434547424, | |
| "learning_rate": 0.0004249306930693069, | |
| "loss": 3.3125, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 14.625, | |
| "grad_norm": 0.3609045445919037, | |
| "learning_rate": 0.0004247559697146185, | |
| "loss": 3.3094, | |
| "step": 50250 | |
| }, | |
| { | |
| "epoch": 14.639555193292967, | |
| "grad_norm": 0.3469443917274475, | |
| "learning_rate": 0.0004245812463599301, | |
| "loss": 3.3116, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 14.654110386585934, | |
| "grad_norm": 0.34178218245506287, | |
| "learning_rate": 0.00042440652300524166, | |
| "loss": 3.3141, | |
| "step": 50350 | |
| }, | |
| { | |
| "epoch": 14.668665579878901, | |
| "grad_norm": 0.3778707683086395, | |
| "learning_rate": 0.00042423179965055325, | |
| "loss": 3.319, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 14.683220773171868, | |
| "grad_norm": 0.3536614179611206, | |
| "learning_rate": 0.00042405707629586484, | |
| "loss": 3.3098, | |
| "step": 50450 | |
| }, | |
| { | |
| "epoch": 14.697775966464835, | |
| "grad_norm": 0.3431127667427063, | |
| "learning_rate": 0.0004238823529411764, | |
| "loss": 3.3091, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 14.712331159757802, | |
| "grad_norm": 0.41096821427345276, | |
| "learning_rate": 0.00042370762958648803, | |
| "loss": 3.3104, | |
| "step": 50550 | |
| }, | |
| { | |
| "epoch": 14.72688635305077, | |
| "grad_norm": 0.3570299446582794, | |
| "learning_rate": 0.00042353290623179963, | |
| "loss": 3.3163, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 14.741441546343735, | |
| "grad_norm": 0.3669699728488922, | |
| "learning_rate": 0.0004233581828771112, | |
| "loss": 3.318, | |
| "step": 50650 | |
| }, | |
| { | |
| "epoch": 14.755996739636702, | |
| "grad_norm": 0.3691817820072174, | |
| "learning_rate": 0.00042318345952242276, | |
| "loss": 3.3236, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 14.770551932929669, | |
| "grad_norm": 0.381858229637146, | |
| "learning_rate": 0.00042300873616773436, | |
| "loss": 3.3118, | |
| "step": 50750 | |
| }, | |
| { | |
| "epoch": 14.785107126222636, | |
| "grad_norm": 0.3354921340942383, | |
| "learning_rate": 0.00042283401281304595, | |
| "loss": 3.3126, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 14.799662319515603, | |
| "grad_norm": 0.377597838640213, | |
| "learning_rate": 0.0004226592894583576, | |
| "loss": 3.3184, | |
| "step": 50850 | |
| }, | |
| { | |
| "epoch": 14.81421751280857, | |
| "grad_norm": 0.3841676414012909, | |
| "learning_rate": 0.00042248456610366914, | |
| "loss": 3.3036, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 14.828772706101537, | |
| "grad_norm": 0.36683332920074463, | |
| "learning_rate": 0.00042230984274898073, | |
| "loss": 3.3153, | |
| "step": 50950 | |
| }, | |
| { | |
| "epoch": 14.843327899394504, | |
| "grad_norm": 0.3898123800754547, | |
| "learning_rate": 0.00042213511939429233, | |
| "loss": 3.3212, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.843327899394504, | |
| "eval_accuracy": 0.3720270439731939, | |
| "eval_loss": 3.5415496826171875, | |
| "eval_runtime": 181.548, | |
| "eval_samples_per_second": 91.717, | |
| "eval_steps_per_second": 5.734, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.857883092687471, | |
| "grad_norm": 0.35734856128692627, | |
| "learning_rate": 0.0004219603960396039, | |
| "loss": 3.318, | |
| "step": 51050 | |
| }, | |
| { | |
| "epoch": 14.872438285980438, | |
| "grad_norm": 0.35610759258270264, | |
| "learning_rate": 0.0004217856726849155, | |
| "loss": 3.3265, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 14.886993479273405, | |
| "grad_norm": 0.37935730814933777, | |
| "learning_rate": 0.0004216109493302271, | |
| "loss": 3.3323, | |
| "step": 51150 | |
| }, | |
| { | |
| "epoch": 14.901548672566372, | |
| "grad_norm": 0.353719562292099, | |
| "learning_rate": 0.0004214362259755387, | |
| "loss": 3.3173, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 14.916103865859338, | |
| "grad_norm": 0.3561287522315979, | |
| "learning_rate": 0.0004212615026208503, | |
| "loss": 3.3289, | |
| "step": 51250 | |
| }, | |
| { | |
| "epoch": 14.930659059152305, | |
| "grad_norm": 0.3487884998321533, | |
| "learning_rate": 0.00042108677926616184, | |
| "loss": 3.3334, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 14.945214252445272, | |
| "grad_norm": 0.3653643727302551, | |
| "learning_rate": 0.00042091205591147343, | |
| "loss": 3.3138, | |
| "step": 51350 | |
| }, | |
| { | |
| "epoch": 14.959769445738239, | |
| "grad_norm": 0.3630554974079132, | |
| "learning_rate": 0.0004207373325567851, | |
| "loss": 3.3223, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 14.974324639031206, | |
| "grad_norm": 0.3576192259788513, | |
| "learning_rate": 0.0004205626092020967, | |
| "loss": 3.3339, | |
| "step": 51450 | |
| }, | |
| { | |
| "epoch": 14.988879832324173, | |
| "grad_norm": 0.3210159242153168, | |
| "learning_rate": 0.0004203878858474082, | |
| "loss": 3.338, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 15.003202142524453, | |
| "grad_norm": 0.3387458920478821, | |
| "learning_rate": 0.0004202131624927198, | |
| "loss": 3.3036, | |
| "step": 51550 | |
| }, | |
| { | |
| "epoch": 15.01775733581742, | |
| "grad_norm": 0.3409118950366974, | |
| "learning_rate": 0.0004200384391380314, | |
| "loss": 3.2119, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 15.032312529110387, | |
| "grad_norm": 0.353676438331604, | |
| "learning_rate": 0.00041986371578334305, | |
| "loss": 3.2213, | |
| "step": 51650 | |
| }, | |
| { | |
| "epoch": 15.046867722403354, | |
| "grad_norm": 0.3806433379650116, | |
| "learning_rate": 0.0004196889924286546, | |
| "loss": 3.2249, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 15.06142291569632, | |
| "grad_norm": 0.3820704519748688, | |
| "learning_rate": 0.0004195142690739662, | |
| "loss": 3.2229, | |
| "step": 51750 | |
| }, | |
| { | |
| "epoch": 15.075978108989288, | |
| "grad_norm": 0.3519267141819, | |
| "learning_rate": 0.0004193395457192778, | |
| "loss": 3.2291, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 15.090533302282255, | |
| "grad_norm": 0.34987378120422363, | |
| "learning_rate": 0.0004191648223645893, | |
| "loss": 3.229, | |
| "step": 51850 | |
| }, | |
| { | |
| "epoch": 15.105088495575222, | |
| "grad_norm": 0.352450430393219, | |
| "learning_rate": 0.0004189900990099009, | |
| "loss": 3.2363, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 15.119643688868187, | |
| "grad_norm": 0.35399603843688965, | |
| "learning_rate": 0.00041881537565521256, | |
| "loss": 3.2366, | |
| "step": 51950 | |
| }, | |
| { | |
| "epoch": 15.134198882161154, | |
| "grad_norm": 0.38175299763679504, | |
| "learning_rate": 0.00041864065230052416, | |
| "loss": 3.2308, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.134198882161154, | |
| "eval_accuracy": 0.3711719152990112, | |
| "eval_loss": 3.5574145317077637, | |
| "eval_runtime": 181.4096, | |
| "eval_samples_per_second": 91.787, | |
| "eval_steps_per_second": 5.738, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.148754075454121, | |
| "grad_norm": 0.36907628178596497, | |
| "learning_rate": 0.00041846592894583575, | |
| "loss": 3.2495, | |
| "step": 52050 | |
| }, | |
| { | |
| "epoch": 15.163309268747089, | |
| "grad_norm": 0.3483155071735382, | |
| "learning_rate": 0.0004182912055911473, | |
| "loss": 3.243, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 15.177864462040056, | |
| "grad_norm": 0.3661825656890869, | |
| "learning_rate": 0.0004181164822364589, | |
| "loss": 3.2469, | |
| "step": 52150 | |
| }, | |
| { | |
| "epoch": 15.192419655333023, | |
| "grad_norm": 0.3702370822429657, | |
| "learning_rate": 0.0004179417588817705, | |
| "loss": 3.2505, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 15.20697484862599, | |
| "grad_norm": 0.3534018099308014, | |
| "learning_rate": 0.00041776703552708213, | |
| "loss": 3.259, | |
| "step": 52250 | |
| }, | |
| { | |
| "epoch": 15.221530041918957, | |
| "grad_norm": 0.3597104549407959, | |
| "learning_rate": 0.00041759231217239367, | |
| "loss": 3.2638, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 15.236085235211924, | |
| "grad_norm": 0.43977323174476624, | |
| "learning_rate": 0.00041741758881770527, | |
| "loss": 3.2585, | |
| "step": 52350 | |
| }, | |
| { | |
| "epoch": 15.25064042850489, | |
| "grad_norm": 0.36657196283340454, | |
| "learning_rate": 0.00041724286546301686, | |
| "loss": 3.2616, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 15.265195621797858, | |
| "grad_norm": 0.34935930371284485, | |
| "learning_rate": 0.0004170681421083284, | |
| "loss": 3.2552, | |
| "step": 52450 | |
| }, | |
| { | |
| "epoch": 15.279750815090825, | |
| "grad_norm": 0.36236581206321716, | |
| "learning_rate": 0.00041689341875364005, | |
| "loss": 3.2619, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 15.294306008383792, | |
| "grad_norm": 0.3543926775455475, | |
| "learning_rate": 0.00041671869539895164, | |
| "loss": 3.2646, | |
| "step": 52550 | |
| }, | |
| { | |
| "epoch": 15.30886120167676, | |
| "grad_norm": 0.36431822180747986, | |
| "learning_rate": 0.00041654397204426324, | |
| "loss": 3.2691, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 15.323416394969724, | |
| "grad_norm": 0.3814825415611267, | |
| "learning_rate": 0.0004163692486895748, | |
| "loss": 3.2755, | |
| "step": 52650 | |
| }, | |
| { | |
| "epoch": 15.337971588262691, | |
| "grad_norm": 0.3722764551639557, | |
| "learning_rate": 0.00041619452533488637, | |
| "loss": 3.2726, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 15.352526781555659, | |
| "grad_norm": 0.366241991519928, | |
| "learning_rate": 0.00041601980198019797, | |
| "loss": 3.271, | |
| "step": 52750 | |
| }, | |
| { | |
| "epoch": 15.367081974848626, | |
| "grad_norm": 0.3627392053604126, | |
| "learning_rate": 0.0004158450786255096, | |
| "loss": 3.2759, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 15.381637168141593, | |
| "grad_norm": 0.3613549768924713, | |
| "learning_rate": 0.00041567035527082115, | |
| "loss": 3.2707, | |
| "step": 52850 | |
| }, | |
| { | |
| "epoch": 15.39619236143456, | |
| "grad_norm": 0.37892386317253113, | |
| "learning_rate": 0.00041549563191613275, | |
| "loss": 3.2789, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 15.410747554727527, | |
| "grad_norm": 0.36703553795814514, | |
| "learning_rate": 0.00041532090856144434, | |
| "loss": 3.2885, | |
| "step": 52950 | |
| }, | |
| { | |
| "epoch": 15.425302748020494, | |
| "grad_norm": 0.3401341736316681, | |
| "learning_rate": 0.00041514618520675594, | |
| "loss": 3.288, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.425302748020494, | |
| "eval_accuracy": 0.3716571855430602, | |
| "eval_loss": 3.5500683784484863, | |
| "eval_runtime": 181.3534, | |
| "eval_samples_per_second": 91.815, | |
| "eval_steps_per_second": 5.74, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.439857941313461, | |
| "grad_norm": 0.4023256003856659, | |
| "learning_rate": 0.00041497146185206753, | |
| "loss": 3.2749, | |
| "step": 53050 | |
| }, | |
| { | |
| "epoch": 15.454413134606428, | |
| "grad_norm": 0.37236660718917847, | |
| "learning_rate": 0.0004147967384973791, | |
| "loss": 3.2778, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 15.468968327899395, | |
| "grad_norm": 0.37774816155433655, | |
| "learning_rate": 0.0004146220151426907, | |
| "loss": 3.2788, | |
| "step": 53150 | |
| }, | |
| { | |
| "epoch": 15.483523521192362, | |
| "grad_norm": 0.3866797983646393, | |
| "learning_rate": 0.0004144472917880023, | |
| "loss": 3.279, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 15.49807871448533, | |
| "grad_norm": 0.3538343012332916, | |
| "learning_rate": 0.00041427256843331385, | |
| "loss": 3.2892, | |
| "step": 53250 | |
| }, | |
| { | |
| "epoch": 15.512633907778294, | |
| "grad_norm": 0.35101500153541565, | |
| "learning_rate": 0.00041409784507862545, | |
| "loss": 3.2939, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 15.527189101071261, | |
| "grad_norm": 0.3746303915977478, | |
| "learning_rate": 0.0004139231217239371, | |
| "loss": 3.2691, | |
| "step": 53350 | |
| }, | |
| { | |
| "epoch": 15.541744294364229, | |
| "grad_norm": 0.33723315596580505, | |
| "learning_rate": 0.0004137483983692487, | |
| "loss": 3.2896, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 15.556299487657196, | |
| "grad_norm": 0.3884109556674957, | |
| "learning_rate": 0.00041357367501456023, | |
| "loss": 3.2849, | |
| "step": 53450 | |
| }, | |
| { | |
| "epoch": 15.570854680950163, | |
| "grad_norm": 0.37353938817977905, | |
| "learning_rate": 0.0004133989516598718, | |
| "loss": 3.2796, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 15.58540987424313, | |
| "grad_norm": 0.38491731882095337, | |
| "learning_rate": 0.0004132242283051834, | |
| "loss": 3.2967, | |
| "step": 53550 | |
| }, | |
| { | |
| "epoch": 15.599965067536097, | |
| "grad_norm": 0.37890195846557617, | |
| "learning_rate": 0.00041304950495049496, | |
| "loss": 3.3009, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 15.614520260829064, | |
| "grad_norm": 0.351624995470047, | |
| "learning_rate": 0.0004128747815958066, | |
| "loss": 3.3001, | |
| "step": 53650 | |
| }, | |
| { | |
| "epoch": 15.629075454122031, | |
| "grad_norm": 0.3783181309700012, | |
| "learning_rate": 0.0004127000582411182, | |
| "loss": 3.2972, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 15.643630647414998, | |
| "grad_norm": 0.35349804162979126, | |
| "learning_rate": 0.0004125253348864298, | |
| "loss": 3.2933, | |
| "step": 53750 | |
| }, | |
| { | |
| "epoch": 15.658185840707965, | |
| "grad_norm": 0.3687611520290375, | |
| "learning_rate": 0.00041235061153174134, | |
| "loss": 3.2993, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 15.672741034000932, | |
| "grad_norm": 0.3746609091758728, | |
| "learning_rate": 0.00041217588817705293, | |
| "loss": 3.2854, | |
| "step": 53850 | |
| }, | |
| { | |
| "epoch": 15.6872962272939, | |
| "grad_norm": 0.3506755828857422, | |
| "learning_rate": 0.0004120011648223646, | |
| "loss": 3.3065, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 15.701851420586866, | |
| "grad_norm": 0.3722805678844452, | |
| "learning_rate": 0.0004118264414676762, | |
| "loss": 3.287, | |
| "step": 53950 | |
| }, | |
| { | |
| "epoch": 15.716406613879832, | |
| "grad_norm": 0.3747727572917938, | |
| "learning_rate": 0.0004116517181129877, | |
| "loss": 3.2921, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.716406613879832, | |
| "eval_accuracy": 0.3723015877586379, | |
| "eval_loss": 3.541686534881592, | |
| "eval_runtime": 181.4595, | |
| "eval_samples_per_second": 91.762, | |
| "eval_steps_per_second": 5.737, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.730961807172799, | |
| "grad_norm": 0.39570698142051697, | |
| "learning_rate": 0.0004114769947582993, | |
| "loss": 3.2945, | |
| "step": 54050 | |
| }, | |
| { | |
| "epoch": 15.745517000465766, | |
| "grad_norm": 0.35266363620758057, | |
| "learning_rate": 0.0004113022714036109, | |
| "loss": 3.3048, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 15.760072193758733, | |
| "grad_norm": 0.38566699624061584, | |
| "learning_rate": 0.0004111275480489225, | |
| "loss": 3.2989, | |
| "step": 54150 | |
| }, | |
| { | |
| "epoch": 15.7746273870517, | |
| "grad_norm": 0.3701910078525543, | |
| "learning_rate": 0.0004109528246942341, | |
| "loss": 3.3014, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 15.789182580344667, | |
| "grad_norm": 0.37288132309913635, | |
| "learning_rate": 0.0004107781013395457, | |
| "loss": 3.3057, | |
| "step": 54250 | |
| }, | |
| { | |
| "epoch": 15.803737773637634, | |
| "grad_norm": 0.34713712334632874, | |
| "learning_rate": 0.0004106033779848573, | |
| "loss": 3.3007, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 15.818292966930601, | |
| "grad_norm": 0.3697455823421478, | |
| "learning_rate": 0.0004104286546301689, | |
| "loss": 3.3109, | |
| "step": 54350 | |
| }, | |
| { | |
| "epoch": 15.832848160223568, | |
| "grad_norm": 0.3423786163330078, | |
| "learning_rate": 0.0004102539312754804, | |
| "loss": 3.2992, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 15.847403353516535, | |
| "grad_norm": 0.3706797957420349, | |
| "learning_rate": 0.00041007920792079206, | |
| "loss": 3.3111, | |
| "step": 54450 | |
| }, | |
| { | |
| "epoch": 15.861958546809502, | |
| "grad_norm": 0.3587998151779175, | |
| "learning_rate": 0.00040990448456610366, | |
| "loss": 3.3116, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 15.87651374010247, | |
| "grad_norm": 0.3819167912006378, | |
| "learning_rate": 0.00040972976121141525, | |
| "loss": 3.3038, | |
| "step": 54550 | |
| }, | |
| { | |
| "epoch": 15.891068933395436, | |
| "grad_norm": 0.37546828389167786, | |
| "learning_rate": 0.0004095550378567268, | |
| "loss": 3.3088, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 15.905624126688402, | |
| "grad_norm": 0.42444273829460144, | |
| "learning_rate": 0.0004093803145020384, | |
| "loss": 3.315, | |
| "step": 54650 | |
| }, | |
| { | |
| "epoch": 15.920179319981369, | |
| "grad_norm": 0.3654751777648926, | |
| "learning_rate": 0.00040920559114735, | |
| "loss": 3.3072, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 15.934734513274336, | |
| "grad_norm": 0.38066214323043823, | |
| "learning_rate": 0.00040903086779266163, | |
| "loss": 3.3202, | |
| "step": 54750 | |
| }, | |
| { | |
| "epoch": 15.949289706567303, | |
| "grad_norm": 0.37775856256484985, | |
| "learning_rate": 0.00040885614443797317, | |
| "loss": 3.3169, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 15.96384489986027, | |
| "grad_norm": 0.37143778800964355, | |
| "learning_rate": 0.00040868142108328476, | |
| "loss": 3.3098, | |
| "step": 54850 | |
| }, | |
| { | |
| "epoch": 15.978400093153237, | |
| "grad_norm": 0.34987694025039673, | |
| "learning_rate": 0.00040850669772859636, | |
| "loss": 3.3059, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 15.992955286446204, | |
| "grad_norm": 0.36554640531539917, | |
| "learning_rate": 0.0004083319743739079, | |
| "loss": 3.3089, | |
| "step": 54950 | |
| }, | |
| { | |
| "epoch": 16.00727759664648, | |
| "grad_norm": 0.354849636554718, | |
| "learning_rate": 0.0004081572510192195, | |
| "loss": 3.2591, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.00727759664648, | |
| "eval_accuracy": 0.3717398072387653, | |
| "eval_loss": 3.5502583980560303, | |
| "eval_runtime": 181.3395, | |
| "eval_samples_per_second": 91.822, | |
| "eval_steps_per_second": 5.741, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.02183278993945, | |
| "grad_norm": 0.371589720249176, | |
| "learning_rate": 0.00040798252766453114, | |
| "loss": 3.2068, | |
| "step": 55050 | |
| }, | |
| { | |
| "epoch": 16.036387983232416, | |
| "grad_norm": 0.3663135766983032, | |
| "learning_rate": 0.00040780780430984273, | |
| "loss": 3.2002, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 16.050943176525383, | |
| "grad_norm": 0.351938396692276, | |
| "learning_rate": 0.0004076330809551543, | |
| "loss": 3.2036, | |
| "step": 55150 | |
| }, | |
| { | |
| "epoch": 16.06549836981835, | |
| "grad_norm": 0.360164076089859, | |
| "learning_rate": 0.00040745835760046587, | |
| "loss": 3.2188, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 16.080053563111317, | |
| "grad_norm": 0.35275858640670776, | |
| "learning_rate": 0.00040728363424577746, | |
| "loss": 3.2085, | |
| "step": 55250 | |
| }, | |
| { | |
| "epoch": 16.094608756404284, | |
| "grad_norm": 0.38582101464271545, | |
| "learning_rate": 0.0004071089108910891, | |
| "loss": 3.2275, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 16.10916394969725, | |
| "grad_norm": 0.37165072560310364, | |
| "learning_rate": 0.0004069341875364007, | |
| "loss": 3.234, | |
| "step": 55350 | |
| }, | |
| { | |
| "epoch": 16.12371914299022, | |
| "grad_norm": 0.37573471665382385, | |
| "learning_rate": 0.00040675946418171225, | |
| "loss": 3.22, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 16.138274336283185, | |
| "grad_norm": 0.36840030550956726, | |
| "learning_rate": 0.00040658474082702384, | |
| "loss": 3.2255, | |
| "step": 55450 | |
| }, | |
| { | |
| "epoch": 16.152829529576152, | |
| "grad_norm": 0.36897239089012146, | |
| "learning_rate": 0.00040641001747233543, | |
| "loss": 3.2238, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 16.16738472286912, | |
| "grad_norm": 0.3556951582431793, | |
| "learning_rate": 0.000406235294117647, | |
| "loss": 3.2226, | |
| "step": 55550 | |
| }, | |
| { | |
| "epoch": 16.181939916162086, | |
| "grad_norm": 0.3834228813648224, | |
| "learning_rate": 0.0004060605707629586, | |
| "loss": 3.2351, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 16.196495109455054, | |
| "grad_norm": 0.43296241760253906, | |
| "learning_rate": 0.0004058858474082702, | |
| "loss": 3.2453, | |
| "step": 55650 | |
| }, | |
| { | |
| "epoch": 16.21105030274802, | |
| "grad_norm": 0.38260507583618164, | |
| "learning_rate": 0.0004057111240535818, | |
| "loss": 3.2455, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 16.225605496040988, | |
| "grad_norm": 0.37897780537605286, | |
| "learning_rate": 0.00040553640069889335, | |
| "loss": 3.2406, | |
| "step": 55750 | |
| }, | |
| { | |
| "epoch": 16.240160689333955, | |
| "grad_norm": 0.33175134658813477, | |
| "learning_rate": 0.00040536167734420495, | |
| "loss": 3.2564, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 16.254715882626922, | |
| "grad_norm": 0.37075909972190857, | |
| "learning_rate": 0.0004051869539895166, | |
| "loss": 3.2521, | |
| "step": 55850 | |
| }, | |
| { | |
| "epoch": 16.26927107591989, | |
| "grad_norm": 0.34951481223106384, | |
| "learning_rate": 0.0004050122306348282, | |
| "loss": 3.2413, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 16.283826269212856, | |
| "grad_norm": 0.3745467960834503, | |
| "learning_rate": 0.00040483750728013973, | |
| "loss": 3.2515, | |
| "step": 55950 | |
| }, | |
| { | |
| "epoch": 16.298381462505823, | |
| "grad_norm": 0.36970576643943787, | |
| "learning_rate": 0.0004046627839254513, | |
| "loss": 3.2485, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.298381462505823, | |
| "eval_accuracy": 0.37181514224153484, | |
| "eval_loss": 3.5516862869262695, | |
| "eval_runtime": 181.2522, | |
| "eval_samples_per_second": 91.866, | |
| "eval_steps_per_second": 5.743, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.31293665579879, | |
| "grad_norm": 0.4097478985786438, | |
| "learning_rate": 0.0004044880605707629, | |
| "loss": 3.2624, | |
| "step": 56050 | |
| }, | |
| { | |
| "epoch": 16.327491849091757, | |
| "grad_norm": 0.3872320353984833, | |
| "learning_rate": 0.00040431333721607446, | |
| "loss": 3.2618, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 16.342047042384724, | |
| "grad_norm": 0.38024234771728516, | |
| "learning_rate": 0.0004041386138613861, | |
| "loss": 3.2555, | |
| "step": 56150 | |
| }, | |
| { | |
| "epoch": 16.35660223567769, | |
| "grad_norm": 0.35469555854797363, | |
| "learning_rate": 0.0004039638905066977, | |
| "loss": 3.2618, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 16.37115742897066, | |
| "grad_norm": 0.37726670503616333, | |
| "learning_rate": 0.0004037891671520093, | |
| "loss": 3.2522, | |
| "step": 56250 | |
| }, | |
| { | |
| "epoch": 16.385712622263625, | |
| "grad_norm": 0.39300525188446045, | |
| "learning_rate": 0.0004036144437973209, | |
| "loss": 3.249, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 16.40026781555659, | |
| "grad_norm": 0.3485693633556366, | |
| "learning_rate": 0.00040343972044263243, | |
| "loss": 3.265, | |
| "step": 56350 | |
| }, | |
| { | |
| "epoch": 16.414823008849556, | |
| "grad_norm": 0.3581498861312866, | |
| "learning_rate": 0.0004032649970879441, | |
| "loss": 3.2626, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 16.429378202142523, | |
| "grad_norm": 0.3752081096172333, | |
| "learning_rate": 0.00040309027373325567, | |
| "loss": 3.2641, | |
| "step": 56450 | |
| }, | |
| { | |
| "epoch": 16.44393339543549, | |
| "grad_norm": 0.392609179019928, | |
| "learning_rate": 0.00040291555037856727, | |
| "loss": 3.2616, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 16.458488588728457, | |
| "grad_norm": 0.41342949867248535, | |
| "learning_rate": 0.0004027408270238788, | |
| "loss": 3.2656, | |
| "step": 56550 | |
| }, | |
| { | |
| "epoch": 16.473043782021424, | |
| "grad_norm": 0.3561092019081116, | |
| "learning_rate": 0.0004025661036691904, | |
| "loss": 3.2831, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 16.48759897531439, | |
| "grad_norm": 0.3804037868976593, | |
| "learning_rate": 0.000402391380314502, | |
| "loss": 3.2761, | |
| "step": 56650 | |
| }, | |
| { | |
| "epoch": 16.50215416860736, | |
| "grad_norm": 0.3736860454082489, | |
| "learning_rate": 0.00040221665695981364, | |
| "loss": 3.2722, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 16.516709361900325, | |
| "grad_norm": 0.39770689606666565, | |
| "learning_rate": 0.0004020419336051252, | |
| "loss": 3.2705, | |
| "step": 56750 | |
| }, | |
| { | |
| "epoch": 16.531264555193292, | |
| "grad_norm": 0.376775860786438, | |
| "learning_rate": 0.0004018672102504368, | |
| "loss": 3.2794, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 16.54581974848626, | |
| "grad_norm": 0.375805526971817, | |
| "learning_rate": 0.00040169248689574837, | |
| "loss": 3.2835, | |
| "step": 56850 | |
| }, | |
| { | |
| "epoch": 16.560374941779227, | |
| "grad_norm": 0.3479491174221039, | |
| "learning_rate": 0.0004015177635410599, | |
| "loss": 3.2802, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 16.574930135072194, | |
| "grad_norm": 0.3600131869316101, | |
| "learning_rate": 0.0004013430401863715, | |
| "loss": 3.2787, | |
| "step": 56950 | |
| }, | |
| { | |
| "epoch": 16.58948532836516, | |
| "grad_norm": 0.38802334666252136, | |
| "learning_rate": 0.00040116831683168315, | |
| "loss": 3.2739, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.58948532836516, | |
| "eval_accuracy": 0.3727245685308182, | |
| "eval_loss": 3.5458953380584717, | |
| "eval_runtime": 181.516, | |
| "eval_samples_per_second": 91.733, | |
| "eval_steps_per_second": 5.735, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.604040521658128, | |
| "grad_norm": 0.39461463689804077, | |
| "learning_rate": 0.00040099359347699475, | |
| "loss": 3.2684, | |
| "step": 57050 | |
| }, | |
| { | |
| "epoch": 16.618595714951095, | |
| "grad_norm": 0.3506292998790741, | |
| "learning_rate": 0.0004008188701223063, | |
| "loss": 3.2691, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 16.633150908244062, | |
| "grad_norm": 0.3793623745441437, | |
| "learning_rate": 0.0004006441467676179, | |
| "loss": 3.2824, | |
| "step": 57150 | |
| }, | |
| { | |
| "epoch": 16.64770610153703, | |
| "grad_norm": 0.4112677574157715, | |
| "learning_rate": 0.0004004694234129295, | |
| "loss": 3.2887, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 16.662261294829996, | |
| "grad_norm": 0.3722013235092163, | |
| "learning_rate": 0.0004002947000582411, | |
| "loss": 3.2864, | |
| "step": 57250 | |
| }, | |
| { | |
| "epoch": 16.676816488122963, | |
| "grad_norm": 0.4040554165840149, | |
| "learning_rate": 0.00040011997670355267, | |
| "loss": 3.2794, | |
| "step": 57300 | |
| }, | |
| { | |
| "epoch": 16.69137168141593, | |
| "grad_norm": 0.3532615005970001, | |
| "learning_rate": 0.00039994525334886426, | |
| "loss": 3.2882, | |
| "step": 57350 | |
| }, | |
| { | |
| "epoch": 16.705926874708897, | |
| "grad_norm": 0.3731229901313782, | |
| "learning_rate": 0.00039977052999417585, | |
| "loss": 3.2935, | |
| "step": 57400 | |
| }, | |
| { | |
| "epoch": 16.720482068001864, | |
| "grad_norm": 0.3732684552669525, | |
| "learning_rate": 0.00039959580663948745, | |
| "loss": 3.2832, | |
| "step": 57450 | |
| }, | |
| { | |
| "epoch": 16.73503726129483, | |
| "grad_norm": 0.3691073954105377, | |
| "learning_rate": 0.000399421083284799, | |
| "loss": 3.2842, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 16.7495924545878, | |
| "grad_norm": 0.3791075646877289, | |
| "learning_rate": 0.00039924635993011064, | |
| "loss": 3.2934, | |
| "step": 57550 | |
| }, | |
| { | |
| "epoch": 16.764147647880765, | |
| "grad_norm": 0.3449566066265106, | |
| "learning_rate": 0.00039907163657542223, | |
| "loss": 3.2974, | |
| "step": 57600 | |
| }, | |
| { | |
| "epoch": 16.778702841173732, | |
| "grad_norm": 0.3655012547969818, | |
| "learning_rate": 0.0003988969132207338, | |
| "loss": 3.2847, | |
| "step": 57650 | |
| }, | |
| { | |
| "epoch": 16.793258034466696, | |
| "grad_norm": 0.36870133876800537, | |
| "learning_rate": 0.00039872218986604537, | |
| "loss": 3.2877, | |
| "step": 57700 | |
| }, | |
| { | |
| "epoch": 16.807813227759663, | |
| "grad_norm": 0.37629836797714233, | |
| "learning_rate": 0.00039854746651135696, | |
| "loss": 3.2832, | |
| "step": 57750 | |
| }, | |
| { | |
| "epoch": 16.82236842105263, | |
| "grad_norm": 0.3698217272758484, | |
| "learning_rate": 0.0003983727431566686, | |
| "loss": 3.2885, | |
| "step": 57800 | |
| }, | |
| { | |
| "epoch": 16.836923614345597, | |
| "grad_norm": 0.3338254988193512, | |
| "learning_rate": 0.0003981980198019802, | |
| "loss": 3.2886, | |
| "step": 57850 | |
| }, | |
| { | |
| "epoch": 16.851478807638564, | |
| "grad_norm": 0.36517441272735596, | |
| "learning_rate": 0.00039802329644729174, | |
| "loss": 3.3047, | |
| "step": 57900 | |
| }, | |
| { | |
| "epoch": 16.86603400093153, | |
| "grad_norm": 0.3859759569168091, | |
| "learning_rate": 0.00039784857309260334, | |
| "loss": 3.2912, | |
| "step": 57950 | |
| }, | |
| { | |
| "epoch": 16.8805891942245, | |
| "grad_norm": 0.37025895714759827, | |
| "learning_rate": 0.00039767384973791493, | |
| "loss": 3.3028, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.8805891942245, | |
| "eval_accuracy": 0.3730944269609519, | |
| "eval_loss": 3.533510208129883, | |
| "eval_runtime": 181.2342, | |
| "eval_samples_per_second": 91.876, | |
| "eval_steps_per_second": 5.744, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.895144387517465, | |
| "grad_norm": 0.37400710582733154, | |
| "learning_rate": 0.00039749912638322647, | |
| "loss": 3.308, | |
| "step": 58050 | |
| }, | |
| { | |
| "epoch": 16.909699580810432, | |
| "grad_norm": 0.36596307158470154, | |
| "learning_rate": 0.0003973244030285381, | |
| "loss": 3.2998, | |
| "step": 58100 | |
| }, | |
| { | |
| "epoch": 16.9242547741034, | |
| "grad_norm": 0.3551575839519501, | |
| "learning_rate": 0.0003971496796738497, | |
| "loss": 3.3001, | |
| "step": 58150 | |
| }, | |
| { | |
| "epoch": 16.938809967396367, | |
| "grad_norm": 0.3748112618923187, | |
| "learning_rate": 0.0003969749563191613, | |
| "loss": 3.3063, | |
| "step": 58200 | |
| }, | |
| { | |
| "epoch": 16.953365160689334, | |
| "grad_norm": 0.36766737699508667, | |
| "learning_rate": 0.00039680023296447285, | |
| "loss": 3.2918, | |
| "step": 58250 | |
| }, | |
| { | |
| "epoch": 16.9679203539823, | |
| "grad_norm": 0.362453818321228, | |
| "learning_rate": 0.00039662550960978444, | |
| "loss": 3.2946, | |
| "step": 58300 | |
| }, | |
| { | |
| "epoch": 16.982475547275268, | |
| "grad_norm": 0.43086424469947815, | |
| "learning_rate": 0.00039645078625509604, | |
| "loss": 3.2994, | |
| "step": 58350 | |
| }, | |
| { | |
| "epoch": 16.997030740568235, | |
| "grad_norm": 0.3485860526561737, | |
| "learning_rate": 0.0003962760629004077, | |
| "loss": 3.2895, | |
| "step": 58400 | |
| }, | |
| { | |
| "epoch": 17.011353050768513, | |
| "grad_norm": 0.3714965581893921, | |
| "learning_rate": 0.0003961013395457193, | |
| "loss": 3.2196, | |
| "step": 58450 | |
| }, | |
| { | |
| "epoch": 17.02590824406148, | |
| "grad_norm": 0.3558444678783417, | |
| "learning_rate": 0.0003959266161910308, | |
| "loss": 3.1859, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 17.040463437354447, | |
| "grad_norm": 0.42480990290641785, | |
| "learning_rate": 0.0003957518928363424, | |
| "loss": 3.1988, | |
| "step": 58550 | |
| }, | |
| { | |
| "epoch": 17.055018630647414, | |
| "grad_norm": 0.36877307295799255, | |
| "learning_rate": 0.000395577169481654, | |
| "loss": 3.1951, | |
| "step": 58600 | |
| }, | |
| { | |
| "epoch": 17.06957382394038, | |
| "grad_norm": 0.35925930738449097, | |
| "learning_rate": 0.00039540244612696566, | |
| "loss": 3.2045, | |
| "step": 58650 | |
| }, | |
| { | |
| "epoch": 17.084129017233348, | |
| "grad_norm": 0.41032111644744873, | |
| "learning_rate": 0.0003952277227722772, | |
| "loss": 3.1916, | |
| "step": 58700 | |
| }, | |
| { | |
| "epoch": 17.098684210526315, | |
| "grad_norm": 0.40329569578170776, | |
| "learning_rate": 0.0003950529994175888, | |
| "loss": 3.2121, | |
| "step": 58750 | |
| }, | |
| { | |
| "epoch": 17.113239403819282, | |
| "grad_norm": 0.38334032893180847, | |
| "learning_rate": 0.0003948782760629004, | |
| "loss": 3.212, | |
| "step": 58800 | |
| }, | |
| { | |
| "epoch": 17.12779459711225, | |
| "grad_norm": 0.3426177501678467, | |
| "learning_rate": 0.0003947035527082119, | |
| "loss": 3.2076, | |
| "step": 58850 | |
| }, | |
| { | |
| "epoch": 17.142349790405216, | |
| "grad_norm": 0.3961309790611267, | |
| "learning_rate": 0.0003945288293535235, | |
| "loss": 3.2136, | |
| "step": 58900 | |
| }, | |
| { | |
| "epoch": 17.156904983698183, | |
| "grad_norm": 0.386750727891922, | |
| "learning_rate": 0.00039435410599883517, | |
| "loss": 3.2282, | |
| "step": 58950 | |
| }, | |
| { | |
| "epoch": 17.17146017699115, | |
| "grad_norm": 0.3924585282802582, | |
| "learning_rate": 0.00039417938264414676, | |
| "loss": 3.2237, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.17146017699115, | |
| "eval_accuracy": 0.3720184644798988, | |
| "eval_loss": 3.5546765327453613, | |
| "eval_runtime": 180.9754, | |
| "eval_samples_per_second": 92.007, | |
| "eval_steps_per_second": 5.752, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.186015370284117, | |
| "grad_norm": 0.361965149641037, | |
| "learning_rate": 0.0003940046592894583, | |
| "loss": 3.2363, | |
| "step": 59050 | |
| }, | |
| { | |
| "epoch": 17.200570563577084, | |
| "grad_norm": 0.4293610453605652, | |
| "learning_rate": 0.0003938299359347699, | |
| "loss": 3.2393, | |
| "step": 59100 | |
| }, | |
| { | |
| "epoch": 17.21512575687005, | |
| "grad_norm": 0.36987578868865967, | |
| "learning_rate": 0.0003936552125800815, | |
| "loss": 3.221, | |
| "step": 59150 | |
| }, | |
| { | |
| "epoch": 17.22968095016302, | |
| "grad_norm": 0.37416261434555054, | |
| "learning_rate": 0.00039348048922539314, | |
| "loss": 3.2321, | |
| "step": 59200 | |
| }, | |
| { | |
| "epoch": 17.244236143455986, | |
| "grad_norm": 0.37772485613822937, | |
| "learning_rate": 0.0003933057658707047, | |
| "loss": 3.2202, | |
| "step": 59250 | |
| }, | |
| { | |
| "epoch": 17.258791336748953, | |
| "grad_norm": 0.37424352765083313, | |
| "learning_rate": 0.0003931310425160163, | |
| "loss": 3.2311, | |
| "step": 59300 | |
| }, | |
| { | |
| "epoch": 17.27334653004192, | |
| "grad_norm": 0.4201411306858063, | |
| "learning_rate": 0.00039295631916132787, | |
| "loss": 3.2314, | |
| "step": 59350 | |
| }, | |
| { | |
| "epoch": 17.287901723334887, | |
| "grad_norm": 0.3957693874835968, | |
| "learning_rate": 0.00039278159580663946, | |
| "loss": 3.2417, | |
| "step": 59400 | |
| }, | |
| { | |
| "epoch": 17.302456916627854, | |
| "grad_norm": 0.38037219643592834, | |
| "learning_rate": 0.000392606872451951, | |
| "loss": 3.2581, | |
| "step": 59450 | |
| }, | |
| { | |
| "epoch": 17.31701210992082, | |
| "grad_norm": 0.3617865741252899, | |
| "learning_rate": 0.00039243214909726265, | |
| "loss": 3.2369, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 17.331567303213788, | |
| "grad_norm": 0.3748098611831665, | |
| "learning_rate": 0.00039225742574257425, | |
| "loss": 3.2427, | |
| "step": 59550 | |
| }, | |
| { | |
| "epoch": 17.346122496506755, | |
| "grad_norm": 0.4334610104560852, | |
| "learning_rate": 0.00039208270238788584, | |
| "loss": 3.2533, | |
| "step": 59600 | |
| }, | |
| { | |
| "epoch": 17.360677689799722, | |
| "grad_norm": 0.38959386944770813, | |
| "learning_rate": 0.0003919079790331974, | |
| "loss": 3.2523, | |
| "step": 59650 | |
| }, | |
| { | |
| "epoch": 17.37523288309269, | |
| "grad_norm": 0.37862473726272583, | |
| "learning_rate": 0.000391733255678509, | |
| "loss": 3.2447, | |
| "step": 59700 | |
| }, | |
| { | |
| "epoch": 17.389788076385653, | |
| "grad_norm": 0.38255590200424194, | |
| "learning_rate": 0.00039155853232382057, | |
| "loss": 3.2471, | |
| "step": 59750 | |
| }, | |
| { | |
| "epoch": 17.40434326967862, | |
| "grad_norm": 0.38841813802719116, | |
| "learning_rate": 0.0003913838089691322, | |
| "loss": 3.2485, | |
| "step": 59800 | |
| }, | |
| { | |
| "epoch": 17.418898462971587, | |
| "grad_norm": 0.3490172028541565, | |
| "learning_rate": 0.00039120908561444376, | |
| "loss": 3.2544, | |
| "step": 59850 | |
| }, | |
| { | |
| "epoch": 17.433453656264554, | |
| "grad_norm": 0.40020763874053955, | |
| "learning_rate": 0.00039103436225975535, | |
| "loss": 3.2498, | |
| "step": 59900 | |
| }, | |
| { | |
| "epoch": 17.44800884955752, | |
| "grad_norm": 0.39921924471855164, | |
| "learning_rate": 0.00039085963890506695, | |
| "loss": 3.2444, | |
| "step": 59950 | |
| }, | |
| { | |
| "epoch": 17.462564042850488, | |
| "grad_norm": 0.3600024878978729, | |
| "learning_rate": 0.0003906849155503785, | |
| "loss": 3.2656, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.462564042850488, | |
| "eval_accuracy": 0.3723973725125493, | |
| "eval_loss": 3.548039436340332, | |
| "eval_runtime": 181.4348, | |
| "eval_samples_per_second": 91.774, | |
| "eval_steps_per_second": 5.738, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.477119236143455, | |
| "grad_norm": 0.38024771213531494, | |
| "learning_rate": 0.00039051019219569014, | |
| "loss": 3.2521, | |
| "step": 60050 | |
| }, | |
| { | |
| "epoch": 17.491674429436422, | |
| "grad_norm": 0.3785065710544586, | |
| "learning_rate": 0.00039033546884100173, | |
| "loss": 3.2533, | |
| "step": 60100 | |
| }, | |
| { | |
| "epoch": 17.50622962272939, | |
| "grad_norm": 0.3751404285430908, | |
| "learning_rate": 0.0003901607454863133, | |
| "loss": 3.2598, | |
| "step": 60150 | |
| }, | |
| { | |
| "epoch": 17.520784816022356, | |
| "grad_norm": 0.3761499524116516, | |
| "learning_rate": 0.00038998602213162486, | |
| "loss": 3.2586, | |
| "step": 60200 | |
| }, | |
| { | |
| "epoch": 17.535340009315323, | |
| "grad_norm": 0.3843153715133667, | |
| "learning_rate": 0.00038981129877693646, | |
| "loss": 3.2766, | |
| "step": 60250 | |
| }, | |
| { | |
| "epoch": 17.54989520260829, | |
| "grad_norm": 0.3554283082485199, | |
| "learning_rate": 0.00038963657542224805, | |
| "loss": 3.2603, | |
| "step": 60300 | |
| }, | |
| { | |
| "epoch": 17.564450395901257, | |
| "grad_norm": 0.3592737019062042, | |
| "learning_rate": 0.0003894618520675597, | |
| "loss": 3.2806, | |
| "step": 60350 | |
| }, | |
| { | |
| "epoch": 17.579005589194225, | |
| "grad_norm": 0.4006965160369873, | |
| "learning_rate": 0.00038928712871287124, | |
| "loss": 3.2658, | |
| "step": 60400 | |
| }, | |
| { | |
| "epoch": 17.59356078248719, | |
| "grad_norm": 0.352741003036499, | |
| "learning_rate": 0.00038911240535818284, | |
| "loss": 3.2588, | |
| "step": 60450 | |
| }, | |
| { | |
| "epoch": 17.60811597578016, | |
| "grad_norm": 0.4061281681060791, | |
| "learning_rate": 0.00038893768200349443, | |
| "loss": 3.266, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 17.622671169073126, | |
| "grad_norm": 0.39381715655326843, | |
| "learning_rate": 0.000388762958648806, | |
| "loss": 3.266, | |
| "step": 60550 | |
| }, | |
| { | |
| "epoch": 17.637226362366093, | |
| "grad_norm": 0.3868790566921234, | |
| "learning_rate": 0.0003885882352941176, | |
| "loss": 3.2583, | |
| "step": 60600 | |
| }, | |
| { | |
| "epoch": 17.65178155565906, | |
| "grad_norm": 0.3857836425304413, | |
| "learning_rate": 0.0003884135119394292, | |
| "loss": 3.2715, | |
| "step": 60650 | |
| }, | |
| { | |
| "epoch": 17.666336748952027, | |
| "grad_norm": 0.3806355893611908, | |
| "learning_rate": 0.0003882387885847408, | |
| "loss": 3.2631, | |
| "step": 60700 | |
| }, | |
| { | |
| "epoch": 17.680891942244994, | |
| "grad_norm": 0.3960031569004059, | |
| "learning_rate": 0.0003880640652300524, | |
| "loss": 3.2652, | |
| "step": 60750 | |
| }, | |
| { | |
| "epoch": 17.69544713553796, | |
| "grad_norm": 0.35661599040031433, | |
| "learning_rate": 0.00038788934187536394, | |
| "loss": 3.2893, | |
| "step": 60800 | |
| }, | |
| { | |
| "epoch": 17.710002328830928, | |
| "grad_norm": 0.4023573100566864, | |
| "learning_rate": 0.00038771461852067554, | |
| "loss": 3.2819, | |
| "step": 60850 | |
| }, | |
| { | |
| "epoch": 17.724557522123895, | |
| "grad_norm": 0.3875479996204376, | |
| "learning_rate": 0.0003875398951659872, | |
| "loss": 3.2738, | |
| "step": 60900 | |
| }, | |
| { | |
| "epoch": 17.739112715416862, | |
| "grad_norm": 0.3642485439777374, | |
| "learning_rate": 0.0003873651718112988, | |
| "loss": 3.2685, | |
| "step": 60950 | |
| }, | |
| { | |
| "epoch": 17.75366790870983, | |
| "grad_norm": 0.3675863742828369, | |
| "learning_rate": 0.0003871904484566103, | |
| "loss": 3.2953, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.75366790870983, | |
| "eval_accuracy": 0.3730756225920859, | |
| "eval_loss": 3.542440176010132, | |
| "eval_runtime": 181.1914, | |
| "eval_samples_per_second": 91.897, | |
| "eval_steps_per_second": 5.745, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.768223102002796, | |
| "grad_norm": 0.3717908561229706, | |
| "learning_rate": 0.0003870157251019219, | |
| "loss": 3.2746, | |
| "step": 61050 | |
| }, | |
| { | |
| "epoch": 17.78277829529576, | |
| "grad_norm": 0.37193652987480164, | |
| "learning_rate": 0.0003868410017472335, | |
| "loss": 3.2775, | |
| "step": 61100 | |
| }, | |
| { | |
| "epoch": 17.797333488588727, | |
| "grad_norm": 0.37539011240005493, | |
| "learning_rate": 0.00038666627839254505, | |
| "loss": 3.264, | |
| "step": 61150 | |
| }, | |
| { | |
| "epoch": 17.811888681881694, | |
| "grad_norm": 0.4052475392818451, | |
| "learning_rate": 0.0003864915550378567, | |
| "loss": 3.2752, | |
| "step": 61200 | |
| }, | |
| { | |
| "epoch": 17.82644387517466, | |
| "grad_norm": 0.37633731961250305, | |
| "learning_rate": 0.0003863168316831683, | |
| "loss": 3.2831, | |
| "step": 61250 | |
| }, | |
| { | |
| "epoch": 17.840999068467628, | |
| "grad_norm": 0.3629589378833771, | |
| "learning_rate": 0.0003861421083284799, | |
| "loss": 3.2811, | |
| "step": 61300 | |
| }, | |
| { | |
| "epoch": 17.855554261760595, | |
| "grad_norm": 0.35611119866371155, | |
| "learning_rate": 0.0003859673849737914, | |
| "loss": 3.2793, | |
| "step": 61350 | |
| }, | |
| { | |
| "epoch": 17.870109455053562, | |
| "grad_norm": 0.3677215278148651, | |
| "learning_rate": 0.000385792661619103, | |
| "loss": 3.2767, | |
| "step": 61400 | |
| }, | |
| { | |
| "epoch": 17.88466464834653, | |
| "grad_norm": 0.34761014580726624, | |
| "learning_rate": 0.00038561793826441467, | |
| "loss": 3.2798, | |
| "step": 61450 | |
| }, | |
| { | |
| "epoch": 17.899219841639496, | |
| "grad_norm": 0.3875502049922943, | |
| "learning_rate": 0.00038544321490972626, | |
| "loss": 3.3052, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 17.913775034932463, | |
| "grad_norm": 0.3841918110847473, | |
| "learning_rate": 0.0003852684915550378, | |
| "loss": 3.2864, | |
| "step": 61550 | |
| }, | |
| { | |
| "epoch": 17.92833022822543, | |
| "grad_norm": 0.4031478762626648, | |
| "learning_rate": 0.0003850937682003494, | |
| "loss": 3.2933, | |
| "step": 61600 | |
| }, | |
| { | |
| "epoch": 17.942885421518397, | |
| "grad_norm": 0.3854524791240692, | |
| "learning_rate": 0.000384919044845661, | |
| "loss": 3.2834, | |
| "step": 61650 | |
| }, | |
| { | |
| "epoch": 17.957440614811365, | |
| "grad_norm": 0.3800252676010132, | |
| "learning_rate": 0.0003847443214909726, | |
| "loss": 3.2959, | |
| "step": 61700 | |
| }, | |
| { | |
| "epoch": 17.97199580810433, | |
| "grad_norm": 0.33671489357948303, | |
| "learning_rate": 0.00038456959813628423, | |
| "loss": 3.292, | |
| "step": 61750 | |
| }, | |
| { | |
| "epoch": 17.9865510013973, | |
| "grad_norm": 0.40121355652809143, | |
| "learning_rate": 0.0003843948747815958, | |
| "loss": 3.2815, | |
| "step": 61800 | |
| }, | |
| { | |
| "epoch": 18.000873311597577, | |
| "grad_norm": 0.41646862030029297, | |
| "learning_rate": 0.00038422015142690737, | |
| "loss": 3.2722, | |
| "step": 61850 | |
| }, | |
| { | |
| "epoch": 18.015428504890544, | |
| "grad_norm": 0.37640494108200073, | |
| "learning_rate": 0.00038404542807221896, | |
| "loss": 3.1641, | |
| "step": 61900 | |
| }, | |
| { | |
| "epoch": 18.02998369818351, | |
| "grad_norm": 0.3735230267047882, | |
| "learning_rate": 0.0003838707047175305, | |
| "loss": 3.185, | |
| "step": 61950 | |
| }, | |
| { | |
| "epoch": 18.044538891476478, | |
| "grad_norm": 0.3715892732143402, | |
| "learning_rate": 0.00038369598136284215, | |
| "loss": 3.1951, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.044538891476478, | |
| "eval_accuracy": 0.3726110371537895, | |
| "eval_loss": 3.549298048019409, | |
| "eval_runtime": 181.408, | |
| "eval_samples_per_second": 91.788, | |
| "eval_steps_per_second": 5.738, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.059094084769445, | |
| "grad_norm": 0.382521390914917, | |
| "learning_rate": 0.00038352125800815374, | |
| "loss": 3.196, | |
| "step": 62050 | |
| }, | |
| { | |
| "epoch": 18.073649278062412, | |
| "grad_norm": 0.3650335371494293, | |
| "learning_rate": 0.00038334653465346534, | |
| "loss": 3.1978, | |
| "step": 62100 | |
| }, | |
| { | |
| "epoch": 18.08820447135538, | |
| "grad_norm": 0.38035500049591064, | |
| "learning_rate": 0.0003831718112987769, | |
| "loss": 3.1877, | |
| "step": 62150 | |
| }, | |
| { | |
| "epoch": 18.102759664648346, | |
| "grad_norm": 0.37595996260643005, | |
| "learning_rate": 0.0003829970879440885, | |
| "loss": 3.1956, | |
| "step": 62200 | |
| }, | |
| { | |
| "epoch": 18.117314857941313, | |
| "grad_norm": 0.36986905336380005, | |
| "learning_rate": 0.00038282236458940007, | |
| "loss": 3.2052, | |
| "step": 62250 | |
| }, | |
| { | |
| "epoch": 18.13187005123428, | |
| "grad_norm": 0.3834410309791565, | |
| "learning_rate": 0.0003826476412347117, | |
| "loss": 3.1984, | |
| "step": 62300 | |
| }, | |
| { | |
| "epoch": 18.146425244527247, | |
| "grad_norm": 0.42468026280403137, | |
| "learning_rate": 0.00038247291788002326, | |
| "loss": 3.2093, | |
| "step": 62350 | |
| }, | |
| { | |
| "epoch": 18.160980437820214, | |
| "grad_norm": 0.3679347336292267, | |
| "learning_rate": 0.00038229819452533485, | |
| "loss": 3.2119, | |
| "step": 62400 | |
| }, | |
| { | |
| "epoch": 18.17553563111318, | |
| "grad_norm": 0.36006471514701843, | |
| "learning_rate": 0.00038212347117064644, | |
| "loss": 3.2029, | |
| "step": 62450 | |
| }, | |
| { | |
| "epoch": 18.19009082440615, | |
| "grad_norm": 0.4088258445262909, | |
| "learning_rate": 0.000381948747815958, | |
| "loss": 3.2286, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 18.204646017699115, | |
| "grad_norm": 0.35326552391052246, | |
| "learning_rate": 0.0003817740244612696, | |
| "loss": 3.2145, | |
| "step": 62550 | |
| }, | |
| { | |
| "epoch": 18.219201210992082, | |
| "grad_norm": 0.4518735408782959, | |
| "learning_rate": 0.00038159930110658123, | |
| "loss": 3.2278, | |
| "step": 62600 | |
| }, | |
| { | |
| "epoch": 18.23375640428505, | |
| "grad_norm": 0.3534115254878998, | |
| "learning_rate": 0.0003814245777518928, | |
| "loss": 3.2222, | |
| "step": 62650 | |
| }, | |
| { | |
| "epoch": 18.248311597578017, | |
| "grad_norm": 0.3919824957847595, | |
| "learning_rate": 0.0003812498543972044, | |
| "loss": 3.2244, | |
| "step": 62700 | |
| }, | |
| { | |
| "epoch": 18.262866790870984, | |
| "grad_norm": 0.38733547925949097, | |
| "learning_rate": 0.00038107513104251596, | |
| "loss": 3.2259, | |
| "step": 62750 | |
| }, | |
| { | |
| "epoch": 18.27742198416395, | |
| "grad_norm": 0.36184003949165344, | |
| "learning_rate": 0.00038090040768782755, | |
| "loss": 3.2251, | |
| "step": 62800 | |
| }, | |
| { | |
| "epoch": 18.291977177456918, | |
| "grad_norm": 0.3423421084880829, | |
| "learning_rate": 0.0003807256843331392, | |
| "loss": 3.2321, | |
| "step": 62850 | |
| }, | |
| { | |
| "epoch": 18.306532370749885, | |
| "grad_norm": 0.40122663974761963, | |
| "learning_rate": 0.0003805509609784508, | |
| "loss": 3.2213, | |
| "step": 62900 | |
| }, | |
| { | |
| "epoch": 18.321087564042852, | |
| "grad_norm": 0.3773971199989319, | |
| "learning_rate": 0.00038037623762376233, | |
| "loss": 3.2357, | |
| "step": 62950 | |
| }, | |
| { | |
| "epoch": 18.33564275733582, | |
| "grad_norm": 0.3741346001625061, | |
| "learning_rate": 0.00038020151426907393, | |
| "loss": 3.2331, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.33564275733582, | |
| "eval_accuracy": 0.3728661889338405, | |
| "eval_loss": 3.5485711097717285, | |
| "eval_runtime": 181.173, | |
| "eval_samples_per_second": 91.907, | |
| "eval_steps_per_second": 5.746, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.350197950628786, | |
| "grad_norm": 0.37724754214286804, | |
| "learning_rate": 0.0003800267909143855, | |
| "loss": 3.239, | |
| "step": 63050 | |
| }, | |
| { | |
| "epoch": 18.364753143921753, | |
| "grad_norm": 0.34363576769828796, | |
| "learning_rate": 0.00037985206755969706, | |
| "loss": 3.2444, | |
| "step": 63100 | |
| }, | |
| { | |
| "epoch": 18.379308337214717, | |
| "grad_norm": 0.3696479797363281, | |
| "learning_rate": 0.0003796773442050087, | |
| "loss": 3.234, | |
| "step": 63150 | |
| }, | |
| { | |
| "epoch": 18.393863530507684, | |
| "grad_norm": 0.3984706997871399, | |
| "learning_rate": 0.0003795026208503203, | |
| "loss": 3.2427, | |
| "step": 63200 | |
| }, | |
| { | |
| "epoch": 18.40841872380065, | |
| "grad_norm": 0.39606693387031555, | |
| "learning_rate": 0.0003793278974956319, | |
| "loss": 3.2468, | |
| "step": 63250 | |
| }, | |
| { | |
| "epoch": 18.422973917093618, | |
| "grad_norm": 0.4036598801612854, | |
| "learning_rate": 0.00037915317414094344, | |
| "loss": 3.2501, | |
| "step": 63300 | |
| }, | |
| { | |
| "epoch": 18.437529110386585, | |
| "grad_norm": 0.34738820791244507, | |
| "learning_rate": 0.00037897845078625503, | |
| "loss": 3.2571, | |
| "step": 63350 | |
| }, | |
| { | |
| "epoch": 18.452084303679552, | |
| "grad_norm": 0.36835139989852905, | |
| "learning_rate": 0.0003788037274315667, | |
| "loss": 3.249, | |
| "step": 63400 | |
| }, | |
| { | |
| "epoch": 18.46663949697252, | |
| "grad_norm": 0.3729427456855774, | |
| "learning_rate": 0.0003786290040768783, | |
| "loss": 3.2576, | |
| "step": 63450 | |
| }, | |
| { | |
| "epoch": 18.481194690265486, | |
| "grad_norm": 0.38652747869491577, | |
| "learning_rate": 0.0003784542807221898, | |
| "loss": 3.2487, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 18.495749883558453, | |
| "grad_norm": 0.3909215033054352, | |
| "learning_rate": 0.0003782795573675014, | |
| "loss": 3.2474, | |
| "step": 63550 | |
| }, | |
| { | |
| "epoch": 18.51030507685142, | |
| "grad_norm": 0.3667212426662445, | |
| "learning_rate": 0.000378104834012813, | |
| "loss": 3.2556, | |
| "step": 63600 | |
| }, | |
| { | |
| "epoch": 18.524860270144387, | |
| "grad_norm": 0.36072465777397156, | |
| "learning_rate": 0.0003779301106581246, | |
| "loss": 3.2449, | |
| "step": 63650 | |
| }, | |
| { | |
| "epoch": 18.539415463437354, | |
| "grad_norm": 0.375143438577652, | |
| "learning_rate": 0.0003777553873034362, | |
| "loss": 3.2497, | |
| "step": 63700 | |
| }, | |
| { | |
| "epoch": 18.55397065673032, | |
| "grad_norm": 0.3688351511955261, | |
| "learning_rate": 0.0003775806639487478, | |
| "loss": 3.2531, | |
| "step": 63750 | |
| }, | |
| { | |
| "epoch": 18.56852585002329, | |
| "grad_norm": 0.3653005361557007, | |
| "learning_rate": 0.0003774059405940594, | |
| "loss": 3.2502, | |
| "step": 63800 | |
| }, | |
| { | |
| "epoch": 18.583081043316255, | |
| "grad_norm": 0.3686894178390503, | |
| "learning_rate": 0.000377231217239371, | |
| "loss": 3.2417, | |
| "step": 63850 | |
| }, | |
| { | |
| "epoch": 18.597636236609222, | |
| "grad_norm": 0.3764590620994568, | |
| "learning_rate": 0.0003770564938846825, | |
| "loss": 3.2539, | |
| "step": 63900 | |
| }, | |
| { | |
| "epoch": 18.61219142990219, | |
| "grad_norm": 0.395633727312088, | |
| "learning_rate": 0.00037688177052999416, | |
| "loss": 3.264, | |
| "step": 63950 | |
| }, | |
| { | |
| "epoch": 18.626746623195157, | |
| "grad_norm": 0.3682929575443268, | |
| "learning_rate": 0.00037670704717530576, | |
| "loss": 3.2593, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.626746623195157, | |
| "eval_accuracy": 0.3727078796534496, | |
| "eval_loss": 3.540853500366211, | |
| "eval_runtime": 181.1321, | |
| "eval_samples_per_second": 91.927, | |
| "eval_steps_per_second": 5.747, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.641301816488124, | |
| "grad_norm": 0.4006809592247009, | |
| "learning_rate": 0.00037653232382061735, | |
| "loss": 3.2549, | |
| "step": 64050 | |
| }, | |
| { | |
| "epoch": 18.65585700978109, | |
| "grad_norm": 0.3779636025428772, | |
| "learning_rate": 0.0003763576004659289, | |
| "loss": 3.2531, | |
| "step": 64100 | |
| }, | |
| { | |
| "epoch": 18.670412203074058, | |
| "grad_norm": 0.3712959289550781, | |
| "learning_rate": 0.0003761828771112405, | |
| "loss": 3.2488, | |
| "step": 64150 | |
| }, | |
| { | |
| "epoch": 18.684967396367025, | |
| "grad_norm": 0.39823856949806213, | |
| "learning_rate": 0.0003760081537565521, | |
| "loss": 3.2561, | |
| "step": 64200 | |
| }, | |
| { | |
| "epoch": 18.699522589659992, | |
| "grad_norm": 0.4071546494960785, | |
| "learning_rate": 0.00037583343040186373, | |
| "loss": 3.2578, | |
| "step": 64250 | |
| }, | |
| { | |
| "epoch": 18.71407778295296, | |
| "grad_norm": 0.3799132704734802, | |
| "learning_rate": 0.00037565870704717527, | |
| "loss": 3.2622, | |
| "step": 64300 | |
| }, | |
| { | |
| "epoch": 18.728632976245926, | |
| "grad_norm": 0.394061416387558, | |
| "learning_rate": 0.00037548398369248687, | |
| "loss": 3.2412, | |
| "step": 64350 | |
| }, | |
| { | |
| "epoch": 18.743188169538893, | |
| "grad_norm": 0.39441555738449097, | |
| "learning_rate": 0.00037530926033779846, | |
| "loss": 3.2708, | |
| "step": 64400 | |
| }, | |
| { | |
| "epoch": 18.75774336283186, | |
| "grad_norm": 0.38645127415657043, | |
| "learning_rate": 0.00037513453698311, | |
| "loss": 3.2749, | |
| "step": 64450 | |
| }, | |
| { | |
| "epoch": 18.772298556124824, | |
| "grad_norm": 0.3764881193637848, | |
| "learning_rate": 0.0003749598136284216, | |
| "loss": 3.2553, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 18.78685374941779, | |
| "grad_norm": 0.3740571439266205, | |
| "learning_rate": 0.00037478509027373324, | |
| "loss": 3.2532, | |
| "step": 64550 | |
| }, | |
| { | |
| "epoch": 18.801408942710758, | |
| "grad_norm": 0.36852291226387024, | |
| "learning_rate": 0.00037461036691904484, | |
| "loss": 3.2661, | |
| "step": 64600 | |
| }, | |
| { | |
| "epoch": 18.815964136003725, | |
| "grad_norm": 0.36674243211746216, | |
| "learning_rate": 0.0003744356435643564, | |
| "loss": 3.2765, | |
| "step": 64650 | |
| }, | |
| { | |
| "epoch": 18.830519329296692, | |
| "grad_norm": 0.3778327703475952, | |
| "learning_rate": 0.00037426092020966797, | |
| "loss": 3.2664, | |
| "step": 64700 | |
| }, | |
| { | |
| "epoch": 18.84507452258966, | |
| "grad_norm": 0.38024887442588806, | |
| "learning_rate": 0.00037408619685497957, | |
| "loss": 3.2679, | |
| "step": 64750 | |
| }, | |
| { | |
| "epoch": 18.859629715882626, | |
| "grad_norm": 0.38589903712272644, | |
| "learning_rate": 0.0003739114735002912, | |
| "loss": 3.2735, | |
| "step": 64800 | |
| }, | |
| { | |
| "epoch": 18.874184909175593, | |
| "grad_norm": 0.3877568542957306, | |
| "learning_rate": 0.0003737367501456028, | |
| "loss": 3.2718, | |
| "step": 64850 | |
| }, | |
| { | |
| "epoch": 18.88874010246856, | |
| "grad_norm": 0.3728450536727905, | |
| "learning_rate": 0.00037356202679091435, | |
| "loss": 3.2666, | |
| "step": 64900 | |
| }, | |
| { | |
| "epoch": 18.903295295761527, | |
| "grad_norm": 0.38396206498146057, | |
| "learning_rate": 0.00037338730343622594, | |
| "loss": 3.2741, | |
| "step": 64950 | |
| }, | |
| { | |
| "epoch": 18.917850489054494, | |
| "grad_norm": 0.37104254961013794, | |
| "learning_rate": 0.00037321258008153754, | |
| "loss": 3.2761, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.917850489054494, | |
| "eval_accuracy": 0.37353950286654974, | |
| "eval_loss": 3.5339014530181885, | |
| "eval_runtime": 181.2275, | |
| "eval_samples_per_second": 91.879, | |
| "eval_steps_per_second": 5.744, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.93240568234746, | |
| "grad_norm": 0.3607959449291229, | |
| "learning_rate": 0.0003730378567268491, | |
| "loss": 3.2735, | |
| "step": 65050 | |
| }, | |
| { | |
| "epoch": 18.94696087564043, | |
| "grad_norm": 0.3835141956806183, | |
| "learning_rate": 0.0003728631333721607, | |
| "loss": 3.2632, | |
| "step": 65100 | |
| }, | |
| { | |
| "epoch": 18.961516068933395, | |
| "grad_norm": 0.4325149357318878, | |
| "learning_rate": 0.0003726884100174723, | |
| "loss": 3.2719, | |
| "step": 65150 | |
| }, | |
| { | |
| "epoch": 18.976071262226363, | |
| "grad_norm": 0.3760964274406433, | |
| "learning_rate": 0.0003725136866627839, | |
| "loss": 3.2746, | |
| "step": 65200 | |
| }, | |
| { | |
| "epoch": 18.99062645551933, | |
| "grad_norm": 0.38402265310287476, | |
| "learning_rate": 0.00037233896330809545, | |
| "loss": 3.2712, | |
| "step": 65250 | |
| }, | |
| { | |
| "epoch": 19.004948765719607, | |
| "grad_norm": 0.38246601819992065, | |
| "learning_rate": 0.00037216423995340705, | |
| "loss": 3.2424, | |
| "step": 65300 | |
| }, | |
| { | |
| "epoch": 19.019503959012575, | |
| "grad_norm": 0.3741970360279083, | |
| "learning_rate": 0.0003719895165987187, | |
| "loss": 3.1709, | |
| "step": 65350 | |
| }, | |
| { | |
| "epoch": 19.03405915230554, | |
| "grad_norm": 0.3912128508090973, | |
| "learning_rate": 0.0003718147932440303, | |
| "loss": 3.1856, | |
| "step": 65400 | |
| }, | |
| { | |
| "epoch": 19.04861434559851, | |
| "grad_norm": 0.3777288496494293, | |
| "learning_rate": 0.00037164006988934183, | |
| "loss": 3.1709, | |
| "step": 65450 | |
| }, | |
| { | |
| "epoch": 19.063169538891476, | |
| "grad_norm": 0.38531720638275146, | |
| "learning_rate": 0.0003714653465346534, | |
| "loss": 3.189, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 19.077724732184443, | |
| "grad_norm": 0.36882486939430237, | |
| "learning_rate": 0.000371290623179965, | |
| "loss": 3.1823, | |
| "step": 65550 | |
| }, | |
| { | |
| "epoch": 19.09227992547741, | |
| "grad_norm": 0.39064404368400574, | |
| "learning_rate": 0.00037111589982527656, | |
| "loss": 3.1747, | |
| "step": 65600 | |
| }, | |
| { | |
| "epoch": 19.106835118770377, | |
| "grad_norm": 0.43773752450942993, | |
| "learning_rate": 0.0003709411764705882, | |
| "loss": 3.1819, | |
| "step": 65650 | |
| }, | |
| { | |
| "epoch": 19.121390312063344, | |
| "grad_norm": 0.4174773693084717, | |
| "learning_rate": 0.0003707664531158998, | |
| "loss": 3.1938, | |
| "step": 65700 | |
| }, | |
| { | |
| "epoch": 19.13594550535631, | |
| "grad_norm": 0.38426661491394043, | |
| "learning_rate": 0.0003705917297612114, | |
| "loss": 3.1937, | |
| "step": 65750 | |
| }, | |
| { | |
| "epoch": 19.150500698649278, | |
| "grad_norm": 0.37709906697273254, | |
| "learning_rate": 0.000370417006406523, | |
| "loss": 3.1885, | |
| "step": 65800 | |
| }, | |
| { | |
| "epoch": 19.165055891942245, | |
| "grad_norm": 0.42608362436294556, | |
| "learning_rate": 0.00037024228305183453, | |
| "loss": 3.1975, | |
| "step": 65850 | |
| }, | |
| { | |
| "epoch": 19.179611085235212, | |
| "grad_norm": 0.4232037365436554, | |
| "learning_rate": 0.0003700675596971461, | |
| "loss": 3.2081, | |
| "step": 65900 | |
| }, | |
| { | |
| "epoch": 19.19416627852818, | |
| "grad_norm": 0.39610755443573, | |
| "learning_rate": 0.0003698928363424578, | |
| "loss": 3.2042, | |
| "step": 65950 | |
| }, | |
| { | |
| "epoch": 19.208721471821146, | |
| "grad_norm": 0.38242438435554504, | |
| "learning_rate": 0.00036971811298776937, | |
| "loss": 3.2115, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.208721471821146, | |
| "eval_accuracy": 0.37263524777870455, | |
| "eval_loss": 3.552818775177002, | |
| "eval_runtime": 181.5094, | |
| "eval_samples_per_second": 91.736, | |
| "eval_steps_per_second": 5.735, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.223276665114113, | |
| "grad_norm": 0.40275952219963074, | |
| "learning_rate": 0.0003695433896330809, | |
| "loss": 3.211, | |
| "step": 66050 | |
| }, | |
| { | |
| "epoch": 19.23783185840708, | |
| "grad_norm": 0.38202306628227234, | |
| "learning_rate": 0.0003693686662783925, | |
| "loss": 3.2086, | |
| "step": 66100 | |
| }, | |
| { | |
| "epoch": 19.252387051700047, | |
| "grad_norm": 0.3808169960975647, | |
| "learning_rate": 0.0003691939429237041, | |
| "loss": 3.2087, | |
| "step": 66150 | |
| }, | |
| { | |
| "epoch": 19.266942244993015, | |
| "grad_norm": 0.3790864646434784, | |
| "learning_rate": 0.00036901921956901575, | |
| "loss": 3.2216, | |
| "step": 66200 | |
| }, | |
| { | |
| "epoch": 19.28149743828598, | |
| "grad_norm": 0.4077747166156769, | |
| "learning_rate": 0.0003688444962143273, | |
| "loss": 3.2075, | |
| "step": 66250 | |
| }, | |
| { | |
| "epoch": 19.29605263157895, | |
| "grad_norm": 0.3944033980369568, | |
| "learning_rate": 0.0003686697728596389, | |
| "loss": 3.2227, | |
| "step": 66300 | |
| }, | |
| { | |
| "epoch": 19.310607824871916, | |
| "grad_norm": 0.3898009955883026, | |
| "learning_rate": 0.0003684950495049505, | |
| "loss": 3.2192, | |
| "step": 66350 | |
| }, | |
| { | |
| "epoch": 19.325163018164883, | |
| "grad_norm": 0.3814161717891693, | |
| "learning_rate": 0.000368320326150262, | |
| "loss": 3.2139, | |
| "step": 66400 | |
| }, | |
| { | |
| "epoch": 19.33971821145785, | |
| "grad_norm": 0.38406553864479065, | |
| "learning_rate": 0.0003681456027955736, | |
| "loss": 3.2201, | |
| "step": 66450 | |
| }, | |
| { | |
| "epoch": 19.354273404750813, | |
| "grad_norm": 0.376544326543808, | |
| "learning_rate": 0.00036797087944088526, | |
| "loss": 3.2216, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 19.36882859804378, | |
| "grad_norm": 0.372443825006485, | |
| "learning_rate": 0.00036779615608619685, | |
| "loss": 3.2304, | |
| "step": 66550 | |
| }, | |
| { | |
| "epoch": 19.383383791336747, | |
| "grad_norm": 0.40152785181999207, | |
| "learning_rate": 0.0003676214327315084, | |
| "loss": 3.23, | |
| "step": 66600 | |
| }, | |
| { | |
| "epoch": 19.397938984629715, | |
| "grad_norm": 0.41309258341789246, | |
| "learning_rate": 0.00036744670937682, | |
| "loss": 3.2264, | |
| "step": 66650 | |
| }, | |
| { | |
| "epoch": 19.41249417792268, | |
| "grad_norm": 0.37425634264945984, | |
| "learning_rate": 0.0003672719860221316, | |
| "loss": 3.2374, | |
| "step": 66700 | |
| }, | |
| { | |
| "epoch": 19.42704937121565, | |
| "grad_norm": 0.3656749725341797, | |
| "learning_rate": 0.00036709726266744323, | |
| "loss": 3.2347, | |
| "step": 66750 | |
| }, | |
| { | |
| "epoch": 19.441604564508616, | |
| "grad_norm": 0.383222371339798, | |
| "learning_rate": 0.00036692253931275477, | |
| "loss": 3.2161, | |
| "step": 66800 | |
| }, | |
| { | |
| "epoch": 19.456159757801583, | |
| "grad_norm": 0.39083653688430786, | |
| "learning_rate": 0.00036674781595806636, | |
| "loss": 3.2301, | |
| "step": 66850 | |
| }, | |
| { | |
| "epoch": 19.47071495109455, | |
| "grad_norm": 0.3562709093093872, | |
| "learning_rate": 0.00036657309260337796, | |
| "loss": 3.2478, | |
| "step": 66900 | |
| }, | |
| { | |
| "epoch": 19.485270144387517, | |
| "grad_norm": 0.34880685806274414, | |
| "learning_rate": 0.00036639836924868955, | |
| "loss": 3.241, | |
| "step": 66950 | |
| }, | |
| { | |
| "epoch": 19.499825337680484, | |
| "grad_norm": 0.40873685479164124, | |
| "learning_rate": 0.0003662236458940011, | |
| "loss": 3.253, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.499825337680484, | |
| "eval_accuracy": 0.373206665537621, | |
| "eval_loss": 3.5417890548706055, | |
| "eval_runtime": 181.1986, | |
| "eval_samples_per_second": 91.894, | |
| "eval_steps_per_second": 5.745, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.51438053097345, | |
| "grad_norm": 0.379344642162323, | |
| "learning_rate": 0.00036604892253931274, | |
| "loss": 3.2529, | |
| "step": 67050 | |
| }, | |
| { | |
| "epoch": 19.528935724266418, | |
| "grad_norm": 0.3883436620235443, | |
| "learning_rate": 0.00036587419918462433, | |
| "loss": 3.2466, | |
| "step": 67100 | |
| }, | |
| { | |
| "epoch": 19.543490917559385, | |
| "grad_norm": 0.3720000088214874, | |
| "learning_rate": 0.00036569947582993593, | |
| "loss": 3.2464, | |
| "step": 67150 | |
| }, | |
| { | |
| "epoch": 19.558046110852352, | |
| "grad_norm": 0.40469080209732056, | |
| "learning_rate": 0.00036552475247524747, | |
| "loss": 3.2474, | |
| "step": 67200 | |
| }, | |
| { | |
| "epoch": 19.57260130414532, | |
| "grad_norm": 0.3750063478946686, | |
| "learning_rate": 0.00036535002912055906, | |
| "loss": 3.2367, | |
| "step": 67250 | |
| }, | |
| { | |
| "epoch": 19.587156497438286, | |
| "grad_norm": 0.38467472791671753, | |
| "learning_rate": 0.00036517530576587066, | |
| "loss": 3.2446, | |
| "step": 67300 | |
| }, | |
| { | |
| "epoch": 19.601711690731253, | |
| "grad_norm": 0.3896279036998749, | |
| "learning_rate": 0.0003650005824111823, | |
| "loss": 3.2494, | |
| "step": 67350 | |
| }, | |
| { | |
| "epoch": 19.61626688402422, | |
| "grad_norm": 0.4054372012615204, | |
| "learning_rate": 0.00036482585905649385, | |
| "loss": 3.2491, | |
| "step": 67400 | |
| }, | |
| { | |
| "epoch": 19.630822077317188, | |
| "grad_norm": 0.3839910328388214, | |
| "learning_rate": 0.00036465113570180544, | |
| "loss": 3.2424, | |
| "step": 67450 | |
| }, | |
| { | |
| "epoch": 19.645377270610155, | |
| "grad_norm": 0.3664180040359497, | |
| "learning_rate": 0.00036447641234711703, | |
| "loss": 3.2461, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 19.65993246390312, | |
| "grad_norm": 0.36931681632995605, | |
| "learning_rate": 0.0003643016889924286, | |
| "loss": 3.2386, | |
| "step": 67550 | |
| }, | |
| { | |
| "epoch": 19.67448765719609, | |
| "grad_norm": 0.3919360935688019, | |
| "learning_rate": 0.0003641269656377402, | |
| "loss": 3.2551, | |
| "step": 67600 | |
| }, | |
| { | |
| "epoch": 19.689042850489056, | |
| "grad_norm": 0.39192062616348267, | |
| "learning_rate": 0.0003639522422830518, | |
| "loss": 3.2434, | |
| "step": 67650 | |
| }, | |
| { | |
| "epoch": 19.703598043782023, | |
| "grad_norm": 0.3523690104484558, | |
| "learning_rate": 0.0003637775189283634, | |
| "loss": 3.2539, | |
| "step": 67700 | |
| }, | |
| { | |
| "epoch": 19.71815323707499, | |
| "grad_norm": 0.3749070167541504, | |
| "learning_rate": 0.00036360279557367495, | |
| "loss": 3.2427, | |
| "step": 67750 | |
| }, | |
| { | |
| "epoch": 19.732708430367957, | |
| "grad_norm": 0.36370980739593506, | |
| "learning_rate": 0.00036342807221898655, | |
| "loss": 3.2471, | |
| "step": 67800 | |
| }, | |
| { | |
| "epoch": 19.74726362366092, | |
| "grad_norm": 0.4108258783817291, | |
| "learning_rate": 0.00036325334886429814, | |
| "loss": 3.2508, | |
| "step": 67850 | |
| }, | |
| { | |
| "epoch": 19.761818816953888, | |
| "grad_norm": 0.4331919848918915, | |
| "learning_rate": 0.0003630786255096098, | |
| "loss": 3.2479, | |
| "step": 67900 | |
| }, | |
| { | |
| "epoch": 19.776374010246855, | |
| "grad_norm": 0.36273193359375, | |
| "learning_rate": 0.00036290390215492133, | |
| "loss": 3.2536, | |
| "step": 67950 | |
| }, | |
| { | |
| "epoch": 19.79092920353982, | |
| "grad_norm": 0.39168423414230347, | |
| "learning_rate": 0.0003627291788002329, | |
| "loss": 3.2466, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.79092920353982, | |
| "eval_accuracy": 0.3734319653820971, | |
| "eval_loss": 3.537184715270996, | |
| "eval_runtime": 181.2507, | |
| "eval_samples_per_second": 91.867, | |
| "eval_steps_per_second": 5.743, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.80548439683279, | |
| "grad_norm": 0.38999927043914795, | |
| "learning_rate": 0.0003625544554455445, | |
| "loss": 3.2529, | |
| "step": 68050 | |
| }, | |
| { | |
| "epoch": 19.820039590125756, | |
| "grad_norm": 0.3893345296382904, | |
| "learning_rate": 0.0003623797320908561, | |
| "loss": 3.253, | |
| "step": 68100 | |
| }, | |
| { | |
| "epoch": 19.834594783418723, | |
| "grad_norm": 0.3700770437717438, | |
| "learning_rate": 0.00036220500873616776, | |
| "loss": 3.2616, | |
| "step": 68150 | |
| }, | |
| { | |
| "epoch": 19.84914997671169, | |
| "grad_norm": 0.4178517162799835, | |
| "learning_rate": 0.0003620302853814793, | |
| "loss": 3.2551, | |
| "step": 68200 | |
| }, | |
| { | |
| "epoch": 19.863705170004657, | |
| "grad_norm": 0.394083708524704, | |
| "learning_rate": 0.0003618555620267909, | |
| "loss": 3.2592, | |
| "step": 68250 | |
| }, | |
| { | |
| "epoch": 19.878260363297624, | |
| "grad_norm": 0.3745860457420349, | |
| "learning_rate": 0.0003616808386721025, | |
| "loss": 3.2504, | |
| "step": 68300 | |
| }, | |
| { | |
| "epoch": 19.89281555659059, | |
| "grad_norm": 0.39990267157554626, | |
| "learning_rate": 0.00036150611531741403, | |
| "loss": 3.2545, | |
| "step": 68350 | |
| }, | |
| { | |
| "epoch": 19.907370749883558, | |
| "grad_norm": 0.37312051653862, | |
| "learning_rate": 0.0003613313919627256, | |
| "loss": 3.2596, | |
| "step": 68400 | |
| }, | |
| { | |
| "epoch": 19.921925943176525, | |
| "grad_norm": 0.38254302740097046, | |
| "learning_rate": 0.00036115666860803727, | |
| "loss": 3.2645, | |
| "step": 68450 | |
| }, | |
| { | |
| "epoch": 19.936481136469492, | |
| "grad_norm": 0.41094058752059937, | |
| "learning_rate": 0.00036098194525334887, | |
| "loss": 3.2695, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 19.95103632976246, | |
| "grad_norm": 0.3820090889930725, | |
| "learning_rate": 0.0003608072218986604, | |
| "loss": 3.2599, | |
| "step": 68550 | |
| }, | |
| { | |
| "epoch": 19.965591523055426, | |
| "grad_norm": 0.3915736675262451, | |
| "learning_rate": 0.000360632498543972, | |
| "loss": 3.2675, | |
| "step": 68600 | |
| }, | |
| { | |
| "epoch": 19.980146716348393, | |
| "grad_norm": 0.3794993758201599, | |
| "learning_rate": 0.0003604577751892836, | |
| "loss": 3.2657, | |
| "step": 68650 | |
| }, | |
| { | |
| "epoch": 19.99470190964136, | |
| "grad_norm": 0.38000932335853577, | |
| "learning_rate": 0.00036028305183459513, | |
| "loss": 3.2654, | |
| "step": 68700 | |
| }, | |
| { | |
| "epoch": 20.00902421984164, | |
| "grad_norm": 0.3856370151042938, | |
| "learning_rate": 0.0003601083284799068, | |
| "loss": 3.1988, | |
| "step": 68750 | |
| }, | |
| { | |
| "epoch": 20.023579413134605, | |
| "grad_norm": 0.39030441641807556, | |
| "learning_rate": 0.0003599336051252184, | |
| "loss": 3.155, | |
| "step": 68800 | |
| }, | |
| { | |
| "epoch": 20.038134606427572, | |
| "grad_norm": 0.38833680748939514, | |
| "learning_rate": 0.00035975888177052997, | |
| "loss": 3.1615, | |
| "step": 68850 | |
| }, | |
| { | |
| "epoch": 20.05268979972054, | |
| "grad_norm": 0.38570043444633484, | |
| "learning_rate": 0.0003595841584158415, | |
| "loss": 3.1685, | |
| "step": 68900 | |
| }, | |
| { | |
| "epoch": 20.067244993013507, | |
| "grad_norm": 0.3800423741340637, | |
| "learning_rate": 0.0003594094350611531, | |
| "loss": 3.1791, | |
| "step": 68950 | |
| }, | |
| { | |
| "epoch": 20.081800186306474, | |
| "grad_norm": 0.4435599148273468, | |
| "learning_rate": 0.00035923471170646475, | |
| "loss": 3.1752, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.081800186306474, | |
| "eval_accuracy": 0.3730468284022598, | |
| "eval_loss": 3.548830986022949, | |
| "eval_runtime": 181.2712, | |
| "eval_samples_per_second": 91.857, | |
| "eval_steps_per_second": 5.743, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.09635537959944, | |
| "grad_norm": 0.40220603346824646, | |
| "learning_rate": 0.00035905998835177635, | |
| "loss": 3.1679, | |
| "step": 69050 | |
| }, | |
| { | |
| "epoch": 20.110910572892408, | |
| "grad_norm": 0.37997275590896606, | |
| "learning_rate": 0.00035888526499708794, | |
| "loss": 3.1812, | |
| "step": 69100 | |
| }, | |
| { | |
| "epoch": 20.125465766185375, | |
| "grad_norm": 0.3896423876285553, | |
| "learning_rate": 0.0003587105416423995, | |
| "loss": 3.1766, | |
| "step": 69150 | |
| }, | |
| { | |
| "epoch": 20.140020959478342, | |
| "grad_norm": 0.38264891505241394, | |
| "learning_rate": 0.0003585358182877111, | |
| "loss": 3.1838, | |
| "step": 69200 | |
| }, | |
| { | |
| "epoch": 20.15457615277131, | |
| "grad_norm": 0.4277355372905731, | |
| "learning_rate": 0.00035836109493302267, | |
| "loss": 3.181, | |
| "step": 69250 | |
| }, | |
| { | |
| "epoch": 20.169131346064276, | |
| "grad_norm": 0.3833529055118561, | |
| "learning_rate": 0.0003581863715783343, | |
| "loss": 3.1973, | |
| "step": 69300 | |
| }, | |
| { | |
| "epoch": 20.183686539357243, | |
| "grad_norm": 0.4084775447845459, | |
| "learning_rate": 0.00035801164822364586, | |
| "loss": 3.1851, | |
| "step": 69350 | |
| }, | |
| { | |
| "epoch": 20.19824173265021, | |
| "grad_norm": 0.3731643855571747, | |
| "learning_rate": 0.00035783692486895745, | |
| "loss": 3.1818, | |
| "step": 69400 | |
| }, | |
| { | |
| "epoch": 20.212796925943177, | |
| "grad_norm": 0.3852198123931885, | |
| "learning_rate": 0.00035766220151426905, | |
| "loss": 3.2066, | |
| "step": 69450 | |
| }, | |
| { | |
| "epoch": 20.227352119236144, | |
| "grad_norm": 0.44227173924446106, | |
| "learning_rate": 0.0003574874781595806, | |
| "loss": 3.2011, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 20.24190731252911, | |
| "grad_norm": 0.3953424394130707, | |
| "learning_rate": 0.00035731275480489224, | |
| "loss": 3.2069, | |
| "step": 69550 | |
| }, | |
| { | |
| "epoch": 20.25646250582208, | |
| "grad_norm": 0.39885902404785156, | |
| "learning_rate": 0.00035713803145020383, | |
| "loss": 3.2066, | |
| "step": 69600 | |
| }, | |
| { | |
| "epoch": 20.271017699115045, | |
| "grad_norm": 0.3794148862361908, | |
| "learning_rate": 0.0003569633080955154, | |
| "loss": 3.2042, | |
| "step": 69650 | |
| }, | |
| { | |
| "epoch": 20.285572892408013, | |
| "grad_norm": 0.3893904983997345, | |
| "learning_rate": 0.00035678858474082697, | |
| "loss": 3.2075, | |
| "step": 69700 | |
| }, | |
| { | |
| "epoch": 20.30012808570098, | |
| "grad_norm": 0.4011684060096741, | |
| "learning_rate": 0.00035661386138613856, | |
| "loss": 3.2043, | |
| "step": 69750 | |
| }, | |
| { | |
| "epoch": 20.314683278993947, | |
| "grad_norm": 0.4057433307170868, | |
| "learning_rate": 0.00035643913803145015, | |
| "loss": 3.2102, | |
| "step": 69800 | |
| }, | |
| { | |
| "epoch": 20.32923847228691, | |
| "grad_norm": 0.36967119574546814, | |
| "learning_rate": 0.0003562644146767618, | |
| "loss": 3.2193, | |
| "step": 69850 | |
| }, | |
| { | |
| "epoch": 20.343793665579877, | |
| "grad_norm": 0.379490464925766, | |
| "learning_rate": 0.00035608969132207334, | |
| "loss": 3.214, | |
| "step": 69900 | |
| }, | |
| { | |
| "epoch": 20.358348858872844, | |
| "grad_norm": 0.4042377769947052, | |
| "learning_rate": 0.00035591496796738494, | |
| "loss": 3.2138, | |
| "step": 69950 | |
| }, | |
| { | |
| "epoch": 20.37290405216581, | |
| "grad_norm": 0.4063800871372223, | |
| "learning_rate": 0.00035574024461269653, | |
| "loss": 3.2125, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.37290405216581, | |
| "eval_accuracy": 0.37332101960578756, | |
| "eval_loss": 3.5475003719329834, | |
| "eval_runtime": 181.2622, | |
| "eval_samples_per_second": 91.861, | |
| "eval_steps_per_second": 5.743, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.38745924545878, | |
| "grad_norm": 0.39343348145484924, | |
| "learning_rate": 0.0003555655212580081, | |
| "loss": 3.224, | |
| "step": 70050 | |
| }, | |
| { | |
| "epoch": 20.402014438751745, | |
| "grad_norm": 0.3925335109233856, | |
| "learning_rate": 0.00035539079790331967, | |
| "loss": 3.2262, | |
| "step": 70100 | |
| }, | |
| { | |
| "epoch": 20.416569632044713, | |
| "grad_norm": 0.3898204565048218, | |
| "learning_rate": 0.0003552160745486313, | |
| "loss": 3.2303, | |
| "step": 70150 | |
| }, | |
| { | |
| "epoch": 20.43112482533768, | |
| "grad_norm": 0.4369237720966339, | |
| "learning_rate": 0.0003550413511939429, | |
| "loss": 3.2281, | |
| "step": 70200 | |
| }, | |
| { | |
| "epoch": 20.445680018630647, | |
| "grad_norm": 0.4013806879520416, | |
| "learning_rate": 0.0003548666278392545, | |
| "loss": 3.2124, | |
| "step": 70250 | |
| }, | |
| { | |
| "epoch": 20.460235211923614, | |
| "grad_norm": 0.4022632837295532, | |
| "learning_rate": 0.00035469190448456604, | |
| "loss": 3.2142, | |
| "step": 70300 | |
| }, | |
| { | |
| "epoch": 20.47479040521658, | |
| "grad_norm": 0.38561418652534485, | |
| "learning_rate": 0.00035451718112987764, | |
| "loss": 3.2286, | |
| "step": 70350 | |
| }, | |
| { | |
| "epoch": 20.489345598509548, | |
| "grad_norm": 0.38620567321777344, | |
| "learning_rate": 0.0003543424577751893, | |
| "loss": 3.2225, | |
| "step": 70400 | |
| }, | |
| { | |
| "epoch": 20.503900791802515, | |
| "grad_norm": 0.38152825832366943, | |
| "learning_rate": 0.0003541677344205009, | |
| "loss": 3.2291, | |
| "step": 70450 | |
| }, | |
| { | |
| "epoch": 20.518455985095482, | |
| "grad_norm": 0.39362338185310364, | |
| "learning_rate": 0.0003539930110658124, | |
| "loss": 3.223, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 20.53301117838845, | |
| "grad_norm": 0.3903370797634125, | |
| "learning_rate": 0.000353818287711124, | |
| "loss": 3.2341, | |
| "step": 70550 | |
| }, | |
| { | |
| "epoch": 20.547566371681416, | |
| "grad_norm": 0.3791417181491852, | |
| "learning_rate": 0.0003536435643564356, | |
| "loss": 3.2477, | |
| "step": 70600 | |
| }, | |
| { | |
| "epoch": 20.562121564974383, | |
| "grad_norm": 0.38322868943214417, | |
| "learning_rate": 0.00035346884100174715, | |
| "loss": 3.2317, | |
| "step": 70650 | |
| }, | |
| { | |
| "epoch": 20.57667675826735, | |
| "grad_norm": 0.36763378977775574, | |
| "learning_rate": 0.0003532941176470588, | |
| "loss": 3.2345, | |
| "step": 70700 | |
| }, | |
| { | |
| "epoch": 20.591231951560317, | |
| "grad_norm": 0.39628368616104126, | |
| "learning_rate": 0.0003531193942923704, | |
| "loss": 3.2353, | |
| "step": 70750 | |
| }, | |
| { | |
| "epoch": 20.605787144853284, | |
| "grad_norm": 0.38605067133903503, | |
| "learning_rate": 0.000352944670937682, | |
| "loss": 3.2283, | |
| "step": 70800 | |
| }, | |
| { | |
| "epoch": 20.62034233814625, | |
| "grad_norm": 0.40509700775146484, | |
| "learning_rate": 0.0003527699475829935, | |
| "loss": 3.24, | |
| "step": 70850 | |
| }, | |
| { | |
| "epoch": 20.63489753143922, | |
| "grad_norm": 0.3870425820350647, | |
| "learning_rate": 0.0003525952242283051, | |
| "loss": 3.2256, | |
| "step": 70900 | |
| }, | |
| { | |
| "epoch": 20.649452724732186, | |
| "grad_norm": 0.3848491907119751, | |
| "learning_rate": 0.00035242050087361677, | |
| "loss": 3.2362, | |
| "step": 70950 | |
| }, | |
| { | |
| "epoch": 20.664007918025153, | |
| "grad_norm": 0.3847205340862274, | |
| "learning_rate": 0.00035224577751892836, | |
| "loss": 3.2385, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.664007918025153, | |
| "eval_accuracy": 0.37386528855715373, | |
| "eval_loss": 3.5357706546783447, | |
| "eval_runtime": 181.0515, | |
| "eval_samples_per_second": 91.968, | |
| "eval_steps_per_second": 5.75, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.67856311131812, | |
| "grad_norm": 0.40746915340423584, | |
| "learning_rate": 0.0003520710541642399, | |
| "loss": 3.2387, | |
| "step": 71050 | |
| }, | |
| { | |
| "epoch": 20.693118304611087, | |
| "grad_norm": 0.39892295002937317, | |
| "learning_rate": 0.0003518963308095515, | |
| "loss": 3.2394, | |
| "step": 71100 | |
| }, | |
| { | |
| "epoch": 20.707673497904054, | |
| "grad_norm": 0.3839132487773895, | |
| "learning_rate": 0.0003517216074548631, | |
| "loss": 3.2403, | |
| "step": 71150 | |
| }, | |
| { | |
| "epoch": 20.722228691197017, | |
| "grad_norm": 0.37013792991638184, | |
| "learning_rate": 0.0003515468841001747, | |
| "loss": 3.224, | |
| "step": 71200 | |
| }, | |
| { | |
| "epoch": 20.736783884489984, | |
| "grad_norm": 0.3782537877559662, | |
| "learning_rate": 0.00035137216074548634, | |
| "loss": 3.2499, | |
| "step": 71250 | |
| }, | |
| { | |
| "epoch": 20.75133907778295, | |
| "grad_norm": 0.40229395031929016, | |
| "learning_rate": 0.0003511974373907979, | |
| "loss": 3.2352, | |
| "step": 71300 | |
| }, | |
| { | |
| "epoch": 20.76589427107592, | |
| "grad_norm": 0.4002828598022461, | |
| "learning_rate": 0.00035102271403610947, | |
| "loss": 3.244, | |
| "step": 71350 | |
| }, | |
| { | |
| "epoch": 20.780449464368886, | |
| "grad_norm": 0.3857050836086273, | |
| "learning_rate": 0.00035084799068142106, | |
| "loss": 3.2383, | |
| "step": 71400 | |
| }, | |
| { | |
| "epoch": 20.795004657661853, | |
| "grad_norm": 0.3786620795726776, | |
| "learning_rate": 0.0003506732673267326, | |
| "loss": 3.2368, | |
| "step": 71450 | |
| }, | |
| { | |
| "epoch": 20.80955985095482, | |
| "grad_norm": 0.38194626569747925, | |
| "learning_rate": 0.00035049854397204425, | |
| "loss": 3.2364, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 20.824115044247787, | |
| "grad_norm": 0.3764079511165619, | |
| "learning_rate": 0.00035032382061735585, | |
| "loss": 3.2616, | |
| "step": 71550 | |
| }, | |
| { | |
| "epoch": 20.838670237540754, | |
| "grad_norm": 0.37126094102859497, | |
| "learning_rate": 0.00035014909726266744, | |
| "loss": 3.2465, | |
| "step": 71600 | |
| }, | |
| { | |
| "epoch": 20.85322543083372, | |
| "grad_norm": 0.38610726594924927, | |
| "learning_rate": 0.000349974373907979, | |
| "loss": 3.2539, | |
| "step": 71650 | |
| }, | |
| { | |
| "epoch": 20.867780624126688, | |
| "grad_norm": 0.4158027470111847, | |
| "learning_rate": 0.0003497996505532906, | |
| "loss": 3.2367, | |
| "step": 71700 | |
| }, | |
| { | |
| "epoch": 20.882335817419655, | |
| "grad_norm": 0.4391123652458191, | |
| "learning_rate": 0.00034962492719860217, | |
| "loss": 3.2588, | |
| "step": 71750 | |
| }, | |
| { | |
| "epoch": 20.896891010712622, | |
| "grad_norm": 0.39286503195762634, | |
| "learning_rate": 0.0003494502038439138, | |
| "loss": 3.2484, | |
| "step": 71800 | |
| }, | |
| { | |
| "epoch": 20.91144620400559, | |
| "grad_norm": 0.400235652923584, | |
| "learning_rate": 0.00034927548048922536, | |
| "loss": 3.2569, | |
| "step": 71850 | |
| }, | |
| { | |
| "epoch": 20.926001397298556, | |
| "grad_norm": 0.382038414478302, | |
| "learning_rate": 0.00034910075713453695, | |
| "loss": 3.2467, | |
| "step": 71900 | |
| }, | |
| { | |
| "epoch": 20.940556590591523, | |
| "grad_norm": 0.38471704721450806, | |
| "learning_rate": 0.00034892603377984855, | |
| "loss": 3.2527, | |
| "step": 71950 | |
| }, | |
| { | |
| "epoch": 20.95511178388449, | |
| "grad_norm": 0.369704931974411, | |
| "learning_rate": 0.0003487513104251601, | |
| "loss": 3.251, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.95511178388449, | |
| "eval_accuracy": 0.3739182933718948, | |
| "eval_loss": 3.5311179161071777, | |
| "eval_runtime": 181.3153, | |
| "eval_samples_per_second": 91.835, | |
| "eval_steps_per_second": 5.741, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.969666977177457, | |
| "grad_norm": 0.41386517882347107, | |
| "learning_rate": 0.0003485765870704717, | |
| "loss": 3.2598, | |
| "step": 72050 | |
| }, | |
| { | |
| "epoch": 20.984222170470424, | |
| "grad_norm": 0.39041802287101746, | |
| "learning_rate": 0.00034840186371578333, | |
| "loss": 3.2575, | |
| "step": 72100 | |
| }, | |
| { | |
| "epoch": 20.99877736376339, | |
| "grad_norm": 0.39289528131484985, | |
| "learning_rate": 0.0003482271403610949, | |
| "loss": 3.2441, | |
| "step": 72150 | |
| }, | |
| { | |
| "epoch": 21.01309967396367, | |
| "grad_norm": 0.408292680978775, | |
| "learning_rate": 0.0003480524170064065, | |
| "loss": 3.1657, | |
| "step": 72200 | |
| }, | |
| { | |
| "epoch": 21.027654867256636, | |
| "grad_norm": 0.3830980956554413, | |
| "learning_rate": 0.00034787769365171806, | |
| "loss": 3.1464, | |
| "step": 72250 | |
| }, | |
| { | |
| "epoch": 21.042210060549603, | |
| "grad_norm": 0.3988492488861084, | |
| "learning_rate": 0.00034770297029702965, | |
| "loss": 3.1504, | |
| "step": 72300 | |
| }, | |
| { | |
| "epoch": 21.05676525384257, | |
| "grad_norm": 0.38097283244132996, | |
| "learning_rate": 0.0003475282469423413, | |
| "loss": 3.1562, | |
| "step": 72350 | |
| }, | |
| { | |
| "epoch": 21.071320447135538, | |
| "grad_norm": 0.38642212748527527, | |
| "learning_rate": 0.0003473535235876529, | |
| "loss": 3.1615, | |
| "step": 72400 | |
| }, | |
| { | |
| "epoch": 21.085875640428505, | |
| "grad_norm": 0.39626067876815796, | |
| "learning_rate": 0.00034717880023296444, | |
| "loss": 3.1706, | |
| "step": 72450 | |
| }, | |
| { | |
| "epoch": 21.10043083372147, | |
| "grad_norm": 0.4080914855003357, | |
| "learning_rate": 0.00034700407687827603, | |
| "loss": 3.1722, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 21.11498602701444, | |
| "grad_norm": 0.38640791177749634, | |
| "learning_rate": 0.0003468293535235876, | |
| "loss": 3.1707, | |
| "step": 72550 | |
| }, | |
| { | |
| "epoch": 21.129541220307406, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00034665463016889916, | |
| "loss": 3.1776, | |
| "step": 72600 | |
| }, | |
| { | |
| "epoch": 21.144096413600373, | |
| "grad_norm": 0.3920632302761078, | |
| "learning_rate": 0.0003464799068142108, | |
| "loss": 3.177, | |
| "step": 72650 | |
| }, | |
| { | |
| "epoch": 21.15865160689334, | |
| "grad_norm": 0.407919317483902, | |
| "learning_rate": 0.0003463051834595224, | |
| "loss": 3.173, | |
| "step": 72700 | |
| }, | |
| { | |
| "epoch": 21.173206800186307, | |
| "grad_norm": 0.3902853727340698, | |
| "learning_rate": 0.000346130460104834, | |
| "loss": 3.1717, | |
| "step": 72750 | |
| }, | |
| { | |
| "epoch": 21.187761993479274, | |
| "grad_norm": 0.38187336921691895, | |
| "learning_rate": 0.00034595573675014554, | |
| "loss": 3.1878, | |
| "step": 72800 | |
| }, | |
| { | |
| "epoch": 21.20231718677224, | |
| "grad_norm": 0.4419068992137909, | |
| "learning_rate": 0.00034578101339545714, | |
| "loss": 3.1885, | |
| "step": 72850 | |
| }, | |
| { | |
| "epoch": 21.216872380065208, | |
| "grad_norm": 0.4039742350578308, | |
| "learning_rate": 0.0003456062900407688, | |
| "loss": 3.1867, | |
| "step": 72900 | |
| }, | |
| { | |
| "epoch": 21.231427573358175, | |
| "grad_norm": 0.3693220317363739, | |
| "learning_rate": 0.0003454315666860804, | |
| "loss": 3.2003, | |
| "step": 72950 | |
| }, | |
| { | |
| "epoch": 21.245982766651142, | |
| "grad_norm": 0.41296273469924927, | |
| "learning_rate": 0.0003452568433313919, | |
| "loss": 3.1939, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.245982766651142, | |
| "eval_accuracy": 0.37289486559636115, | |
| "eval_loss": 3.548574447631836, | |
| "eval_runtime": 181.1623, | |
| "eval_samples_per_second": 91.912, | |
| "eval_steps_per_second": 5.746, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.26053795994411, | |
| "grad_norm": 0.38118070363998413, | |
| "learning_rate": 0.0003450821199767035, | |
| "loss": 3.1947, | |
| "step": 73050 | |
| }, | |
| { | |
| "epoch": 21.275093153237076, | |
| "grad_norm": 0.3997765779495239, | |
| "learning_rate": 0.0003449073966220151, | |
| "loss": 3.206, | |
| "step": 73100 | |
| }, | |
| { | |
| "epoch": 21.289648346530043, | |
| "grad_norm": 0.4266320466995239, | |
| "learning_rate": 0.0003447326732673267, | |
| "loss": 3.1859, | |
| "step": 73150 | |
| }, | |
| { | |
| "epoch": 21.30420353982301, | |
| "grad_norm": 0.3923620581626892, | |
| "learning_rate": 0.0003445579499126383, | |
| "loss": 3.2073, | |
| "step": 73200 | |
| }, | |
| { | |
| "epoch": 21.318758733115978, | |
| "grad_norm": 0.38114598393440247, | |
| "learning_rate": 0.0003443832265579499, | |
| "loss": 3.1972, | |
| "step": 73250 | |
| }, | |
| { | |
| "epoch": 21.33331392640894, | |
| "grad_norm": 0.3909752368927002, | |
| "learning_rate": 0.0003442085032032615, | |
| "loss": 3.2098, | |
| "step": 73300 | |
| }, | |
| { | |
| "epoch": 21.347869119701908, | |
| "grad_norm": 0.37029755115509033, | |
| "learning_rate": 0.0003440337798485731, | |
| "loss": 3.2042, | |
| "step": 73350 | |
| }, | |
| { | |
| "epoch": 21.362424312994875, | |
| "grad_norm": 0.3990894556045532, | |
| "learning_rate": 0.0003438590564938846, | |
| "loss": 3.2002, | |
| "step": 73400 | |
| }, | |
| { | |
| "epoch": 21.376979506287842, | |
| "grad_norm": 0.4198392927646637, | |
| "learning_rate": 0.0003436843331391962, | |
| "loss": 3.1933, | |
| "step": 73450 | |
| }, | |
| { | |
| "epoch": 21.39153469958081, | |
| "grad_norm": 0.3842645287513733, | |
| "learning_rate": 0.00034350960978450786, | |
| "loss": 3.2007, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 21.406089892873776, | |
| "grad_norm": 0.39544111490249634, | |
| "learning_rate": 0.00034333488642981946, | |
| "loss": 3.2013, | |
| "step": 73550 | |
| }, | |
| { | |
| "epoch": 21.420645086166743, | |
| "grad_norm": 0.42064350843429565, | |
| "learning_rate": 0.000343160163075131, | |
| "loss": 3.2129, | |
| "step": 73600 | |
| }, | |
| { | |
| "epoch": 21.43520027945971, | |
| "grad_norm": 0.38665804266929626, | |
| "learning_rate": 0.0003429854397204426, | |
| "loss": 3.2248, | |
| "step": 73650 | |
| }, | |
| { | |
| "epoch": 21.449755472752678, | |
| "grad_norm": 0.3785528838634491, | |
| "learning_rate": 0.0003428107163657542, | |
| "loss": 3.2113, | |
| "step": 73700 | |
| }, | |
| { | |
| "epoch": 21.464310666045645, | |
| "grad_norm": 0.3859052360057831, | |
| "learning_rate": 0.00034263599301106583, | |
| "loss": 3.203, | |
| "step": 73750 | |
| }, | |
| { | |
| "epoch": 21.47886585933861, | |
| "grad_norm": 0.4086976647377014, | |
| "learning_rate": 0.0003424612696563774, | |
| "loss": 3.2091, | |
| "step": 73800 | |
| }, | |
| { | |
| "epoch": 21.49342105263158, | |
| "grad_norm": 0.41609588265419006, | |
| "learning_rate": 0.00034228654630168897, | |
| "loss": 3.2185, | |
| "step": 73850 | |
| }, | |
| { | |
| "epoch": 21.507976245924546, | |
| "grad_norm": 0.37916097044944763, | |
| "learning_rate": 0.00034211182294700056, | |
| "loss": 3.211, | |
| "step": 73900 | |
| }, | |
| { | |
| "epoch": 21.522531439217513, | |
| "grad_norm": 0.3910274803638458, | |
| "learning_rate": 0.0003419370995923121, | |
| "loss": 3.2232, | |
| "step": 73950 | |
| }, | |
| { | |
| "epoch": 21.53708663251048, | |
| "grad_norm": 0.41739487648010254, | |
| "learning_rate": 0.0003417623762376237, | |
| "loss": 3.2151, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.53708663251048, | |
| "eval_accuracy": 0.3736458650779482, | |
| "eval_loss": 3.5441434383392334, | |
| "eval_runtime": 181.3399, | |
| "eval_samples_per_second": 91.822, | |
| "eval_steps_per_second": 5.741, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.551641825803447, | |
| "grad_norm": 0.40151894092559814, | |
| "learning_rate": 0.00034158765288293534, | |
| "loss": 3.2244, | |
| "step": 74050 | |
| }, | |
| { | |
| "epoch": 21.566197019096414, | |
| "grad_norm": 0.38283640146255493, | |
| "learning_rate": 0.00034141292952824694, | |
| "loss": 3.2235, | |
| "step": 74100 | |
| }, | |
| { | |
| "epoch": 21.58075221238938, | |
| "grad_norm": 0.38787081837654114, | |
| "learning_rate": 0.0003412382061735585, | |
| "loss": 3.2332, | |
| "step": 74150 | |
| }, | |
| { | |
| "epoch": 21.595307405682348, | |
| "grad_norm": 0.43580567836761475, | |
| "learning_rate": 0.0003410634828188701, | |
| "loss": 3.2224, | |
| "step": 74200 | |
| }, | |
| { | |
| "epoch": 21.609862598975315, | |
| "grad_norm": 0.40886545181274414, | |
| "learning_rate": 0.00034088875946418167, | |
| "loss": 3.2235, | |
| "step": 74250 | |
| }, | |
| { | |
| "epoch": 21.624417792268282, | |
| "grad_norm": 0.37370070815086365, | |
| "learning_rate": 0.0003407140361094933, | |
| "loss": 3.2321, | |
| "step": 74300 | |
| }, | |
| { | |
| "epoch": 21.63897298556125, | |
| "grad_norm": 0.3833332061767578, | |
| "learning_rate": 0.00034053931275480486, | |
| "loss": 3.2174, | |
| "step": 74350 | |
| }, | |
| { | |
| "epoch": 21.653528178854216, | |
| "grad_norm": 0.40175867080688477, | |
| "learning_rate": 0.00034036458940011645, | |
| "loss": 3.2303, | |
| "step": 74400 | |
| }, | |
| { | |
| "epoch": 21.668083372147183, | |
| "grad_norm": 0.43237751722335815, | |
| "learning_rate": 0.00034018986604542804, | |
| "loss": 3.2193, | |
| "step": 74450 | |
| }, | |
| { | |
| "epoch": 21.68263856544015, | |
| "grad_norm": 0.38853147625923157, | |
| "learning_rate": 0.00034001514269073964, | |
| "loss": 3.2311, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 21.697193758733118, | |
| "grad_norm": 0.39983823895454407, | |
| "learning_rate": 0.0003398404193360512, | |
| "loss": 3.2339, | |
| "step": 74550 | |
| }, | |
| { | |
| "epoch": 21.711748952026085, | |
| "grad_norm": 0.39615949988365173, | |
| "learning_rate": 0.00033966569598136283, | |
| "loss": 3.234, | |
| "step": 74600 | |
| }, | |
| { | |
| "epoch": 21.726304145319048, | |
| "grad_norm": 0.40111032128334045, | |
| "learning_rate": 0.0003394909726266744, | |
| "loss": 3.2341, | |
| "step": 74650 | |
| }, | |
| { | |
| "epoch": 21.740859338612015, | |
| "grad_norm": 0.3891691565513611, | |
| "learning_rate": 0.000339316249271986, | |
| "loss": 3.2286, | |
| "step": 74700 | |
| }, | |
| { | |
| "epoch": 21.755414531904982, | |
| "grad_norm": 0.37335634231567383, | |
| "learning_rate": 0.00033914152591729756, | |
| "loss": 3.2335, | |
| "step": 74750 | |
| }, | |
| { | |
| "epoch": 21.76996972519795, | |
| "grad_norm": 0.3801760971546173, | |
| "learning_rate": 0.00033896680256260915, | |
| "loss": 3.2348, | |
| "step": 74800 | |
| }, | |
| { | |
| "epoch": 21.784524918490916, | |
| "grad_norm": 0.4102132022380829, | |
| "learning_rate": 0.00033879207920792074, | |
| "loss": 3.2321, | |
| "step": 74850 | |
| }, | |
| { | |
| "epoch": 21.799080111783883, | |
| "grad_norm": 0.40376681089401245, | |
| "learning_rate": 0.0003386173558532324, | |
| "loss": 3.2337, | |
| "step": 74900 | |
| }, | |
| { | |
| "epoch": 21.81363530507685, | |
| "grad_norm": 0.3670161962509155, | |
| "learning_rate": 0.00033844263249854393, | |
| "loss": 3.2325, | |
| "step": 74950 | |
| }, | |
| { | |
| "epoch": 21.828190498369818, | |
| "grad_norm": 0.4087563455104828, | |
| "learning_rate": 0.00033826790914385553, | |
| "loss": 3.2295, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.828190498369818, | |
| "eval_accuracy": 0.37407201908737464, | |
| "eval_loss": 3.5371601581573486, | |
| "eval_runtime": 180.9832, | |
| "eval_samples_per_second": 92.003, | |
| "eval_steps_per_second": 5.752, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.842745691662785, | |
| "grad_norm": 0.37676188349723816, | |
| "learning_rate": 0.0003380931857891671, | |
| "loss": 3.2364, | |
| "step": 75050 | |
| }, | |
| { | |
| "epoch": 21.85730088495575, | |
| "grad_norm": 0.36632487177848816, | |
| "learning_rate": 0.00033791846243447866, | |
| "loss": 3.2412, | |
| "step": 75100 | |
| }, | |
| { | |
| "epoch": 21.87185607824872, | |
| "grad_norm": 0.4055189788341522, | |
| "learning_rate": 0.0003377437390797903, | |
| "loss": 3.2476, | |
| "step": 75150 | |
| }, | |
| { | |
| "epoch": 21.886411271541686, | |
| "grad_norm": 0.4044891595840454, | |
| "learning_rate": 0.0003375690157251019, | |
| "loss": 3.2275, | |
| "step": 75200 | |
| }, | |
| { | |
| "epoch": 21.900966464834653, | |
| "grad_norm": 0.38007423281669617, | |
| "learning_rate": 0.0003373942923704135, | |
| "loss": 3.2427, | |
| "step": 75250 | |
| }, | |
| { | |
| "epoch": 21.91552165812762, | |
| "grad_norm": 0.4170713424682617, | |
| "learning_rate": 0.00033721956901572504, | |
| "loss": 3.2442, | |
| "step": 75300 | |
| }, | |
| { | |
| "epoch": 21.930076851420587, | |
| "grad_norm": 0.3902817964553833, | |
| "learning_rate": 0.00033704484566103663, | |
| "loss": 3.2375, | |
| "step": 75350 | |
| }, | |
| { | |
| "epoch": 21.944632044713554, | |
| "grad_norm": 0.40278515219688416, | |
| "learning_rate": 0.00033687012230634823, | |
| "loss": 3.2522, | |
| "step": 75400 | |
| }, | |
| { | |
| "epoch": 21.95918723800652, | |
| "grad_norm": 0.36589857935905457, | |
| "learning_rate": 0.0003366953989516599, | |
| "loss": 3.2372, | |
| "step": 75450 | |
| }, | |
| { | |
| "epoch": 21.97374243129949, | |
| "grad_norm": 0.403898686170578, | |
| "learning_rate": 0.00033652067559697147, | |
| "loss": 3.2346, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 21.988297624592455, | |
| "grad_norm": 0.4070855975151062, | |
| "learning_rate": 0.000336345952242283, | |
| "loss": 3.2396, | |
| "step": 75550 | |
| }, | |
| { | |
| "epoch": 22.002619934792733, | |
| "grad_norm": 0.3859020173549652, | |
| "learning_rate": 0.0003361712288875946, | |
| "loss": 3.2177, | |
| "step": 75600 | |
| }, | |
| { | |
| "epoch": 22.0171751280857, | |
| "grad_norm": 0.3986937999725342, | |
| "learning_rate": 0.0003359965055329062, | |
| "loss": 3.1351, | |
| "step": 75650 | |
| }, | |
| { | |
| "epoch": 22.031730321378667, | |
| "grad_norm": 0.39367422461509705, | |
| "learning_rate": 0.00033582178217821785, | |
| "loss": 3.1475, | |
| "step": 75700 | |
| }, | |
| { | |
| "epoch": 22.046285514671634, | |
| "grad_norm": 0.40651074051856995, | |
| "learning_rate": 0.0003356470588235294, | |
| "loss": 3.149, | |
| "step": 75750 | |
| }, | |
| { | |
| "epoch": 22.0608407079646, | |
| "grad_norm": 0.38006383180618286, | |
| "learning_rate": 0.000335472335468841, | |
| "loss": 3.1562, | |
| "step": 75800 | |
| }, | |
| { | |
| "epoch": 22.07539590125757, | |
| "grad_norm": 0.40139079093933105, | |
| "learning_rate": 0.0003352976121141526, | |
| "loss": 3.1548, | |
| "step": 75850 | |
| }, | |
| { | |
| "epoch": 22.089951094550536, | |
| "grad_norm": 0.3921520411968231, | |
| "learning_rate": 0.0003351228887594641, | |
| "loss": 3.1596, | |
| "step": 75900 | |
| }, | |
| { | |
| "epoch": 22.104506287843503, | |
| "grad_norm": 0.4345844089984894, | |
| "learning_rate": 0.0003349481654047757, | |
| "loss": 3.1602, | |
| "step": 75950 | |
| }, | |
| { | |
| "epoch": 22.11906148113647, | |
| "grad_norm": 0.38527101278305054, | |
| "learning_rate": 0.00033477344205008736, | |
| "loss": 3.1566, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.11906148113647, | |
| "eval_accuracy": 0.373341939466151, | |
| "eval_loss": 3.554121732711792, | |
| "eval_runtime": 182.5103, | |
| "eval_samples_per_second": 91.233, | |
| "eval_steps_per_second": 5.704, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.133616674429437, | |
| "grad_norm": 0.4101022183895111, | |
| "learning_rate": 0.00033459871869539895, | |
| "loss": 3.167, | |
| "step": 76050 | |
| }, | |
| { | |
| "epoch": 22.148171867722404, | |
| "grad_norm": 0.4076803922653198, | |
| "learning_rate": 0.0003344239953407105, | |
| "loss": 3.1731, | |
| "step": 76100 | |
| }, | |
| { | |
| "epoch": 22.16272706101537, | |
| "grad_norm": 0.4150278866291046, | |
| "learning_rate": 0.0003342492719860221, | |
| "loss": 3.168, | |
| "step": 76150 | |
| }, | |
| { | |
| "epoch": 22.177282254308338, | |
| "grad_norm": 0.39549461007118225, | |
| "learning_rate": 0.0003340745486313337, | |
| "loss": 3.1759, | |
| "step": 76200 | |
| }, | |
| { | |
| "epoch": 22.191837447601305, | |
| "grad_norm": 0.43638280034065247, | |
| "learning_rate": 0.0003338998252766453, | |
| "loss": 3.1879, | |
| "step": 76250 | |
| }, | |
| { | |
| "epoch": 22.206392640894272, | |
| "grad_norm": 0.4284520745277405, | |
| "learning_rate": 0.00033372510192195687, | |
| "loss": 3.1696, | |
| "step": 76300 | |
| }, | |
| { | |
| "epoch": 22.22094783418724, | |
| "grad_norm": 0.39070242643356323, | |
| "learning_rate": 0.00033355037856726847, | |
| "loss": 3.1726, | |
| "step": 76350 | |
| }, | |
| { | |
| "epoch": 22.235503027480206, | |
| "grad_norm": 0.4118767976760864, | |
| "learning_rate": 0.00033337565521258006, | |
| "loss": 3.1793, | |
| "step": 76400 | |
| }, | |
| { | |
| "epoch": 22.250058220773173, | |
| "grad_norm": 0.41845229268074036, | |
| "learning_rate": 0.00033320093185789165, | |
| "loss": 3.1821, | |
| "step": 76450 | |
| }, | |
| { | |
| "epoch": 22.26461341406614, | |
| "grad_norm": 0.39468759298324585, | |
| "learning_rate": 0.0003330262085032032, | |
| "loss": 3.1824, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 22.279168607359107, | |
| "grad_norm": 0.45352470874786377, | |
| "learning_rate": 0.00033285148514851484, | |
| "loss": 3.1823, | |
| "step": 76550 | |
| }, | |
| { | |
| "epoch": 22.293723800652074, | |
| "grad_norm": 0.39214208722114563, | |
| "learning_rate": 0.00033267676179382644, | |
| "loss": 3.1936, | |
| "step": 76600 | |
| }, | |
| { | |
| "epoch": 22.308278993945038, | |
| "grad_norm": 0.39688727259635925, | |
| "learning_rate": 0.00033250203843913803, | |
| "loss": 3.1897, | |
| "step": 76650 | |
| }, | |
| { | |
| "epoch": 22.322834187238005, | |
| "grad_norm": 0.3960036337375641, | |
| "learning_rate": 0.00033232731508444957, | |
| "loss": 3.1958, | |
| "step": 76700 | |
| }, | |
| { | |
| "epoch": 22.337389380530972, | |
| "grad_norm": 0.42960265278816223, | |
| "learning_rate": 0.00033215259172976117, | |
| "loss": 3.1911, | |
| "step": 76750 | |
| }, | |
| { | |
| "epoch": 22.35194457382394, | |
| "grad_norm": 0.39636436104774475, | |
| "learning_rate": 0.00033197786837507276, | |
| "loss": 3.1924, | |
| "step": 76800 | |
| }, | |
| { | |
| "epoch": 22.366499767116906, | |
| "grad_norm": 0.3948790729045868, | |
| "learning_rate": 0.0003318031450203844, | |
| "loss": 3.1953, | |
| "step": 76850 | |
| }, | |
| { | |
| "epoch": 22.381054960409873, | |
| "grad_norm": 0.3913504183292389, | |
| "learning_rate": 0.00033162842166569595, | |
| "loss": 3.1863, | |
| "step": 76900 | |
| }, | |
| { | |
| "epoch": 22.39561015370284, | |
| "grad_norm": 0.3883304297924042, | |
| "learning_rate": 0.00033145369831100754, | |
| "loss": 3.2014, | |
| "step": 76950 | |
| }, | |
| { | |
| "epoch": 22.410165346995807, | |
| "grad_norm": 0.39272409677505493, | |
| "learning_rate": 0.00033127897495631914, | |
| "loss": 3.2004, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.410165346995807, | |
| "eval_accuracy": 0.3737291919374858, | |
| "eval_loss": 3.545900583267212, | |
| "eval_runtime": 181.6134, | |
| "eval_samples_per_second": 91.684, | |
| "eval_steps_per_second": 5.732, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.424720540288774, | |
| "grad_norm": 0.4007713794708252, | |
| "learning_rate": 0.0003311042516016307, | |
| "loss": 3.2017, | |
| "step": 77050 | |
| }, | |
| { | |
| "epoch": 22.43927573358174, | |
| "grad_norm": 0.4593736529350281, | |
| "learning_rate": 0.0003309295282469423, | |
| "loss": 3.1982, | |
| "step": 77100 | |
| }, | |
| { | |
| "epoch": 22.45383092687471, | |
| "grad_norm": 0.3807480037212372, | |
| "learning_rate": 0.0003307548048922539, | |
| "loss": 3.2012, | |
| "step": 77150 | |
| }, | |
| { | |
| "epoch": 22.468386120167676, | |
| "grad_norm": 0.433830201625824, | |
| "learning_rate": 0.0003305800815375655, | |
| "loss": 3.2042, | |
| "step": 77200 | |
| }, | |
| { | |
| "epoch": 22.482941313460643, | |
| "grad_norm": 0.3865396976470947, | |
| "learning_rate": 0.00033040535818287705, | |
| "loss": 3.2004, | |
| "step": 77250 | |
| }, | |
| { | |
| "epoch": 22.49749650675361, | |
| "grad_norm": 0.42019572854042053, | |
| "learning_rate": 0.00033023063482818865, | |
| "loss": 3.2034, | |
| "step": 77300 | |
| }, | |
| { | |
| "epoch": 22.512051700046577, | |
| "grad_norm": 0.43494096398353577, | |
| "learning_rate": 0.00033005591147350024, | |
| "loss": 3.2009, | |
| "step": 77350 | |
| }, | |
| { | |
| "epoch": 22.526606893339544, | |
| "grad_norm": 0.4004029929637909, | |
| "learning_rate": 0.0003298811881188119, | |
| "loss": 3.2162, | |
| "step": 77400 | |
| }, | |
| { | |
| "epoch": 22.54116208663251, | |
| "grad_norm": 0.41305696964263916, | |
| "learning_rate": 0.00032970646476412343, | |
| "loss": 3.1953, | |
| "step": 77450 | |
| }, | |
| { | |
| "epoch": 22.555717279925478, | |
| "grad_norm": 0.4445006549358368, | |
| "learning_rate": 0.000329531741409435, | |
| "loss": 3.2142, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 22.570272473218445, | |
| "grad_norm": 0.40516790747642517, | |
| "learning_rate": 0.0003293570180547466, | |
| "loss": 3.2041, | |
| "step": 77550 | |
| }, | |
| { | |
| "epoch": 22.584827666511412, | |
| "grad_norm": 0.43369901180267334, | |
| "learning_rate": 0.0003291822947000582, | |
| "loss": 3.2064, | |
| "step": 77600 | |
| }, | |
| { | |
| "epoch": 22.59938285980438, | |
| "grad_norm": 0.4414738714694977, | |
| "learning_rate": 0.00032900757134536975, | |
| "loss": 3.2142, | |
| "step": 77650 | |
| }, | |
| { | |
| "epoch": 22.613938053097346, | |
| "grad_norm": 0.41799449920654297, | |
| "learning_rate": 0.0003288328479906814, | |
| "loss": 3.2172, | |
| "step": 77700 | |
| }, | |
| { | |
| "epoch": 22.628493246390313, | |
| "grad_norm": 0.3902125656604767, | |
| "learning_rate": 0.000328658124635993, | |
| "loss": 3.2141, | |
| "step": 77750 | |
| }, | |
| { | |
| "epoch": 22.64304843968328, | |
| "grad_norm": 0.4232015013694763, | |
| "learning_rate": 0.0003284834012813046, | |
| "loss": 3.2053, | |
| "step": 77800 | |
| }, | |
| { | |
| "epoch": 22.657603632976247, | |
| "grad_norm": 0.3856189250946045, | |
| "learning_rate": 0.00032830867792661613, | |
| "loss": 3.2154, | |
| "step": 77850 | |
| }, | |
| { | |
| "epoch": 22.672158826269214, | |
| "grad_norm": 0.4207291007041931, | |
| "learning_rate": 0.0003281339545719277, | |
| "loss": 3.2233, | |
| "step": 77900 | |
| }, | |
| { | |
| "epoch": 22.68671401956218, | |
| "grad_norm": 0.4035766124725342, | |
| "learning_rate": 0.0003279592312172394, | |
| "loss": 3.2174, | |
| "step": 77950 | |
| }, | |
| { | |
| "epoch": 22.701269212855145, | |
| "grad_norm": 0.3936694860458374, | |
| "learning_rate": 0.00032778450786255097, | |
| "loss": 3.2214, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.701269212855145, | |
| "eval_accuracy": 0.37412055786451004, | |
| "eval_loss": 3.540379047393799, | |
| "eval_runtime": 183.076, | |
| "eval_samples_per_second": 90.951, | |
| "eval_steps_per_second": 5.686, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.715824406148112, | |
| "grad_norm": 0.39343857765197754, | |
| "learning_rate": 0.0003276097845078625, | |
| "loss": 3.2047, | |
| "step": 78050 | |
| }, | |
| { | |
| "epoch": 22.73037959944108, | |
| "grad_norm": 0.43328389525413513, | |
| "learning_rate": 0.0003274350611531741, | |
| "loss": 3.2277, | |
| "step": 78100 | |
| }, | |
| { | |
| "epoch": 22.744934792734046, | |
| "grad_norm": 0.39509138464927673, | |
| "learning_rate": 0.0003272603377984857, | |
| "loss": 3.2203, | |
| "step": 78150 | |
| }, | |
| { | |
| "epoch": 22.759489986027013, | |
| "grad_norm": 0.4134383499622345, | |
| "learning_rate": 0.00032708561444379724, | |
| "loss": 3.2174, | |
| "step": 78200 | |
| }, | |
| { | |
| "epoch": 22.77404517931998, | |
| "grad_norm": 0.4262104332447052, | |
| "learning_rate": 0.0003269108910891089, | |
| "loss": 3.2249, | |
| "step": 78250 | |
| }, | |
| { | |
| "epoch": 22.788600372612947, | |
| "grad_norm": 0.3714088201522827, | |
| "learning_rate": 0.0003267361677344205, | |
| "loss": 3.2223, | |
| "step": 78300 | |
| }, | |
| { | |
| "epoch": 22.803155565905914, | |
| "grad_norm": 0.38759660720825195, | |
| "learning_rate": 0.0003265614443797321, | |
| "loss": 3.2291, | |
| "step": 78350 | |
| }, | |
| { | |
| "epoch": 22.81771075919888, | |
| "grad_norm": 0.3645568788051605, | |
| "learning_rate": 0.0003263867210250436, | |
| "loss": 3.2222, | |
| "step": 78400 | |
| }, | |
| { | |
| "epoch": 22.83226595249185, | |
| "grad_norm": 0.3883237838745117, | |
| "learning_rate": 0.0003262119976703552, | |
| "loss": 3.2259, | |
| "step": 78450 | |
| }, | |
| { | |
| "epoch": 22.846821145784816, | |
| "grad_norm": 0.4034235179424286, | |
| "learning_rate": 0.00032603727431566686, | |
| "loss": 3.2299, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 22.861376339077783, | |
| "grad_norm": 0.42304471135139465, | |
| "learning_rate": 0.00032586255096097845, | |
| "loss": 3.2408, | |
| "step": 78550 | |
| }, | |
| { | |
| "epoch": 22.87593153237075, | |
| "grad_norm": 0.3993287980556488, | |
| "learning_rate": 0.00032568782760629005, | |
| "loss": 3.2307, | |
| "step": 78600 | |
| }, | |
| { | |
| "epoch": 22.890486725663717, | |
| "grad_norm": 0.40124648809432983, | |
| "learning_rate": 0.0003255131042516016, | |
| "loss": 3.2391, | |
| "step": 78650 | |
| }, | |
| { | |
| "epoch": 22.905041918956684, | |
| "grad_norm": 0.39223048090934753, | |
| "learning_rate": 0.0003253383808969132, | |
| "loss": 3.2249, | |
| "step": 78700 | |
| }, | |
| { | |
| "epoch": 22.91959711224965, | |
| "grad_norm": 0.38387811183929443, | |
| "learning_rate": 0.0003251636575422248, | |
| "loss": 3.2355, | |
| "step": 78750 | |
| }, | |
| { | |
| "epoch": 22.934152305542618, | |
| "grad_norm": 0.4031982421875, | |
| "learning_rate": 0.0003249889341875364, | |
| "loss": 3.2232, | |
| "step": 78800 | |
| }, | |
| { | |
| "epoch": 22.948707498835585, | |
| "grad_norm": 0.396810382604599, | |
| "learning_rate": 0.00032481421083284796, | |
| "loss": 3.2344, | |
| "step": 78850 | |
| }, | |
| { | |
| "epoch": 22.963262692128552, | |
| "grad_norm": 0.4282434582710266, | |
| "learning_rate": 0.00032463948747815956, | |
| "loss": 3.2466, | |
| "step": 78900 | |
| }, | |
| { | |
| "epoch": 22.97781788542152, | |
| "grad_norm": 0.4010005593299866, | |
| "learning_rate": 0.00032446476412347115, | |
| "loss": 3.2302, | |
| "step": 78950 | |
| }, | |
| { | |
| "epoch": 22.992373078714486, | |
| "grad_norm": 0.39370250701904297, | |
| "learning_rate": 0.0003242900407687827, | |
| "loss": 3.2437, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 22.992373078714486, | |
| "eval_accuracy": 0.3744414074082867, | |
| "eval_loss": 3.5298874378204346, | |
| "eval_runtime": 181.1381, | |
| "eval_samples_per_second": 91.924, | |
| "eval_steps_per_second": 5.747, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 23.006695388914764, | |
| "grad_norm": 0.4299910068511963, | |
| "learning_rate": 0.00032411531741409434, | |
| "loss": 3.188, | |
| "step": 79050 | |
| }, | |
| { | |
| "epoch": 23.02125058220773, | |
| "grad_norm": 0.3970222771167755, | |
| "learning_rate": 0.00032394059405940593, | |
| "loss": 3.1256, | |
| "step": 79100 | |
| }, | |
| { | |
| "epoch": 23.035805775500698, | |
| "grad_norm": 0.43236055970191956, | |
| "learning_rate": 0.00032376587070471753, | |
| "loss": 3.1269, | |
| "step": 79150 | |
| }, | |
| { | |
| "epoch": 23.050360968793665, | |
| "grad_norm": 0.4045545756816864, | |
| "learning_rate": 0.00032359114735002907, | |
| "loss": 3.1355, | |
| "step": 79200 | |
| }, | |
| { | |
| "epoch": 23.064916162086632, | |
| "grad_norm": 0.4147092401981354, | |
| "learning_rate": 0.00032341642399534066, | |
| "loss": 3.1496, | |
| "step": 79250 | |
| }, | |
| { | |
| "epoch": 23.0794713553796, | |
| "grad_norm": 0.4010353088378906, | |
| "learning_rate": 0.00032324170064065226, | |
| "loss": 3.1275, | |
| "step": 79300 | |
| }, | |
| { | |
| "epoch": 23.094026548672566, | |
| "grad_norm": 0.4142921566963196, | |
| "learning_rate": 0.0003230669772859639, | |
| "loss": 3.1604, | |
| "step": 79350 | |
| }, | |
| { | |
| "epoch": 23.108581741965533, | |
| "grad_norm": 0.3941037058830261, | |
| "learning_rate": 0.00032289225393127545, | |
| "loss": 3.1492, | |
| "step": 79400 | |
| }, | |
| { | |
| "epoch": 23.1231369352585, | |
| "grad_norm": 0.39037835597991943, | |
| "learning_rate": 0.00032271753057658704, | |
| "loss": 3.148, | |
| "step": 79450 | |
| }, | |
| { | |
| "epoch": 23.137692128551468, | |
| "grad_norm": 0.4245337247848511, | |
| "learning_rate": 0.00032254280722189863, | |
| "loss": 3.1634, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 23.152247321844435, | |
| "grad_norm": 0.4353491961956024, | |
| "learning_rate": 0.00032236808386721023, | |
| "loss": 3.1494, | |
| "step": 79550 | |
| }, | |
| { | |
| "epoch": 23.1668025151374, | |
| "grad_norm": 0.41144099831581116, | |
| "learning_rate": 0.00032219336051252177, | |
| "loss": 3.1536, | |
| "step": 79600 | |
| }, | |
| { | |
| "epoch": 23.18135770843037, | |
| "grad_norm": 0.4307765066623688, | |
| "learning_rate": 0.0003220186371578334, | |
| "loss": 3.1573, | |
| "step": 79650 | |
| }, | |
| { | |
| "epoch": 23.195912901723336, | |
| "grad_norm": 0.42108792066574097, | |
| "learning_rate": 0.000321843913803145, | |
| "loss": 3.1741, | |
| "step": 79700 | |
| }, | |
| { | |
| "epoch": 23.210468095016303, | |
| "grad_norm": 0.4174800515174866, | |
| "learning_rate": 0.0003216691904484566, | |
| "loss": 3.1674, | |
| "step": 79750 | |
| }, | |
| { | |
| "epoch": 23.22502328830927, | |
| "grad_norm": 0.41323181986808777, | |
| "learning_rate": 0.00032149446709376815, | |
| "loss": 3.1739, | |
| "step": 79800 | |
| }, | |
| { | |
| "epoch": 23.239578481602237, | |
| "grad_norm": 0.40551289916038513, | |
| "learning_rate": 0.00032131974373907974, | |
| "loss": 3.1751, | |
| "step": 79850 | |
| }, | |
| { | |
| "epoch": 23.254133674895204, | |
| "grad_norm": 0.42636001110076904, | |
| "learning_rate": 0.0003211450203843914, | |
| "loss": 3.1804, | |
| "step": 79900 | |
| }, | |
| { | |
| "epoch": 23.26868886818817, | |
| "grad_norm": 0.3733903467655182, | |
| "learning_rate": 0.000320970297029703, | |
| "loss": 3.1768, | |
| "step": 79950 | |
| }, | |
| { | |
| "epoch": 23.283244061481135, | |
| "grad_norm": 0.4172973334789276, | |
| "learning_rate": 0.0003207955736750145, | |
| "loss": 3.1845, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.283244061481135, | |
| "eval_accuracy": 0.37387151750434056, | |
| "eval_loss": 3.5492820739746094, | |
| "eval_runtime": 182.8566, | |
| "eval_samples_per_second": 91.06, | |
| "eval_steps_per_second": 5.693, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.2977992547741, | |
| "grad_norm": 0.41692233085632324, | |
| "learning_rate": 0.0003206208503203261, | |
| "loss": 3.1795, | |
| "step": 80050 | |
| }, | |
| { | |
| "epoch": 23.31235444806707, | |
| "grad_norm": 0.4022136330604553, | |
| "learning_rate": 0.0003204461269656377, | |
| "loss": 3.1751, | |
| "step": 80100 | |
| }, | |
| { | |
| "epoch": 23.326909641360036, | |
| "grad_norm": 0.4068368971347809, | |
| "learning_rate": 0.00032027140361094925, | |
| "loss": 3.1923, | |
| "step": 80150 | |
| }, | |
| { | |
| "epoch": 23.341464834653003, | |
| "grad_norm": 0.40505388379096985, | |
| "learning_rate": 0.0003200966802562609, | |
| "loss": 3.1891, | |
| "step": 80200 | |
| }, | |
| { | |
| "epoch": 23.35602002794597, | |
| "grad_norm": 0.40903159976005554, | |
| "learning_rate": 0.0003199219569015725, | |
| "loss": 3.1887, | |
| "step": 80250 | |
| }, | |
| { | |
| "epoch": 23.370575221238937, | |
| "grad_norm": 0.4178510904312134, | |
| "learning_rate": 0.0003197472335468841, | |
| "loss": 3.1795, | |
| "step": 80300 | |
| }, | |
| { | |
| "epoch": 23.385130414531904, | |
| "grad_norm": 0.401119589805603, | |
| "learning_rate": 0.00031957251019219563, | |
| "loss": 3.1897, | |
| "step": 80350 | |
| }, | |
| { | |
| "epoch": 23.39968560782487, | |
| "grad_norm": 0.3909188508987427, | |
| "learning_rate": 0.0003193977868375072, | |
| "loss": 3.1857, | |
| "step": 80400 | |
| }, | |
| { | |
| "epoch": 23.41424080111784, | |
| "grad_norm": 0.4655745327472687, | |
| "learning_rate": 0.00031922306348281887, | |
| "loss": 3.1892, | |
| "step": 80450 | |
| }, | |
| { | |
| "epoch": 23.428795994410805, | |
| "grad_norm": 0.4173809587955475, | |
| "learning_rate": 0.00031904834012813047, | |
| "loss": 3.1838, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 23.443351187703772, | |
| "grad_norm": 0.4490557312965393, | |
| "learning_rate": 0.000318873616773442, | |
| "loss": 3.2014, | |
| "step": 80550 | |
| }, | |
| { | |
| "epoch": 23.45790638099674, | |
| "grad_norm": 0.4582597315311432, | |
| "learning_rate": 0.0003186988934187536, | |
| "loss": 3.1982, | |
| "step": 80600 | |
| }, | |
| { | |
| "epoch": 23.472461574289706, | |
| "grad_norm": 0.3948287069797516, | |
| "learning_rate": 0.0003185241700640652, | |
| "loss": 3.1967, | |
| "step": 80650 | |
| }, | |
| { | |
| "epoch": 23.487016767582674, | |
| "grad_norm": 0.41583526134490967, | |
| "learning_rate": 0.0003183494467093768, | |
| "loss": 3.2026, | |
| "step": 80700 | |
| }, | |
| { | |
| "epoch": 23.50157196087564, | |
| "grad_norm": 0.41763588786125183, | |
| "learning_rate": 0.0003181747233546884, | |
| "loss": 3.2021, | |
| "step": 80750 | |
| }, | |
| { | |
| "epoch": 23.516127154168608, | |
| "grad_norm": 0.4259377121925354, | |
| "learning_rate": 0.000318, | |
| "loss": 3.2006, | |
| "step": 80800 | |
| }, | |
| { | |
| "epoch": 23.530682347461575, | |
| "grad_norm": 0.409658282995224, | |
| "learning_rate": 0.00031782527664531157, | |
| "loss": 3.198, | |
| "step": 80850 | |
| }, | |
| { | |
| "epoch": 23.545237540754542, | |
| "grad_norm": 0.43902572989463806, | |
| "learning_rate": 0.00031765055329062317, | |
| "loss": 3.2013, | |
| "step": 80900 | |
| }, | |
| { | |
| "epoch": 23.55979273404751, | |
| "grad_norm": 0.40293756127357483, | |
| "learning_rate": 0.0003174758299359347, | |
| "loss": 3.2025, | |
| "step": 80950 | |
| }, | |
| { | |
| "epoch": 23.574347927340476, | |
| "grad_norm": 0.4030076861381531, | |
| "learning_rate": 0.0003173011065812463, | |
| "loss": 3.1994, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.574347927340476, | |
| "eval_accuracy": 0.37436407444132513, | |
| "eval_loss": 3.5441434383392334, | |
| "eval_runtime": 181.3124, | |
| "eval_samples_per_second": 91.836, | |
| "eval_steps_per_second": 5.741, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.588903120633443, | |
| "grad_norm": 0.4462394118309021, | |
| "learning_rate": 0.00031712638322655795, | |
| "loss": 3.1878, | |
| "step": 81050 | |
| }, | |
| { | |
| "epoch": 23.60345831392641, | |
| "grad_norm": 0.42426666617393494, | |
| "learning_rate": 0.00031695165987186954, | |
| "loss": 3.1984, | |
| "step": 81100 | |
| }, | |
| { | |
| "epoch": 23.618013507219377, | |
| "grad_norm": 0.384657084941864, | |
| "learning_rate": 0.0003167769365171811, | |
| "loss": 3.2021, | |
| "step": 81150 | |
| }, | |
| { | |
| "epoch": 23.632568700512344, | |
| "grad_norm": 0.3903048634529114, | |
| "learning_rate": 0.0003166022131624927, | |
| "loss": 3.205, | |
| "step": 81200 | |
| }, | |
| { | |
| "epoch": 23.64712389380531, | |
| "grad_norm": 0.4122599959373474, | |
| "learning_rate": 0.00031642748980780427, | |
| "loss": 3.2009, | |
| "step": 81250 | |
| }, | |
| { | |
| "epoch": 23.66167908709828, | |
| "grad_norm": 0.43961748480796814, | |
| "learning_rate": 0.0003162527664531159, | |
| "loss": 3.204, | |
| "step": 81300 | |
| }, | |
| { | |
| "epoch": 23.676234280391242, | |
| "grad_norm": 0.4120798110961914, | |
| "learning_rate": 0.00031607804309842746, | |
| "loss": 3.1975, | |
| "step": 81350 | |
| }, | |
| { | |
| "epoch": 23.69078947368421, | |
| "grad_norm": 0.40398430824279785, | |
| "learning_rate": 0.00031590331974373905, | |
| "loss": 3.2111, | |
| "step": 81400 | |
| }, | |
| { | |
| "epoch": 23.705344666977176, | |
| "grad_norm": 0.42327627539634705, | |
| "learning_rate": 0.00031572859638905065, | |
| "loss": 3.2023, | |
| "step": 81450 | |
| }, | |
| { | |
| "epoch": 23.719899860270143, | |
| "grad_norm": 0.4177756905555725, | |
| "learning_rate": 0.0003155538730343622, | |
| "loss": 3.2132, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 23.73445505356311, | |
| "grad_norm": 0.4342160224914551, | |
| "learning_rate": 0.0003153791496796738, | |
| "loss": 3.224, | |
| "step": 81550 | |
| }, | |
| { | |
| "epoch": 23.749010246856077, | |
| "grad_norm": 0.4283663034439087, | |
| "learning_rate": 0.00031520442632498543, | |
| "loss": 3.2156, | |
| "step": 81600 | |
| }, | |
| { | |
| "epoch": 23.763565440149044, | |
| "grad_norm": 0.39514467120170593, | |
| "learning_rate": 0.000315029702970297, | |
| "loss": 3.2086, | |
| "step": 81650 | |
| }, | |
| { | |
| "epoch": 23.77812063344201, | |
| "grad_norm": 0.4161839187145233, | |
| "learning_rate": 0.0003148549796156086, | |
| "loss": 3.2062, | |
| "step": 81700 | |
| }, | |
| { | |
| "epoch": 23.79267582673498, | |
| "grad_norm": 0.44791948795318604, | |
| "learning_rate": 0.00031468025626092016, | |
| "loss": 3.212, | |
| "step": 81750 | |
| }, | |
| { | |
| "epoch": 23.807231020027945, | |
| "grad_norm": 0.41878649592399597, | |
| "learning_rate": 0.00031450553290623175, | |
| "loss": 3.2154, | |
| "step": 81800 | |
| }, | |
| { | |
| "epoch": 23.821786213320912, | |
| "grad_norm": 0.39344221353530884, | |
| "learning_rate": 0.0003143308095515434, | |
| "loss": 3.2188, | |
| "step": 81850 | |
| }, | |
| { | |
| "epoch": 23.83634140661388, | |
| "grad_norm": 0.4187600910663605, | |
| "learning_rate": 0.000314156086196855, | |
| "loss": 3.2193, | |
| "step": 81900 | |
| }, | |
| { | |
| "epoch": 23.850896599906847, | |
| "grad_norm": 0.41973283886909485, | |
| "learning_rate": 0.00031398136284216654, | |
| "loss": 3.2182, | |
| "step": 81950 | |
| }, | |
| { | |
| "epoch": 23.865451793199814, | |
| "grad_norm": 0.4089304208755493, | |
| "learning_rate": 0.00031380663948747813, | |
| "loss": 3.2281, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.865451793199814, | |
| "eval_accuracy": 0.37461628803874075, | |
| "eval_loss": 3.532421588897705, | |
| "eval_runtime": 181.294, | |
| "eval_samples_per_second": 91.845, | |
| "eval_steps_per_second": 5.742, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.88000698649278, | |
| "grad_norm": 0.43405380845069885, | |
| "learning_rate": 0.0003136319161327897, | |
| "loss": 3.2204, | |
| "step": 82050 | |
| }, | |
| { | |
| "epoch": 23.894562179785748, | |
| "grad_norm": 0.38422340154647827, | |
| "learning_rate": 0.00031345719277810127, | |
| "loss": 3.2313, | |
| "step": 82100 | |
| }, | |
| { | |
| "epoch": 23.909117373078715, | |
| "grad_norm": 0.4141247272491455, | |
| "learning_rate": 0.0003132824694234129, | |
| "loss": 3.2268, | |
| "step": 82150 | |
| }, | |
| { | |
| "epoch": 23.923672566371682, | |
| "grad_norm": 0.40371742844581604, | |
| "learning_rate": 0.0003131077460687245, | |
| "loss": 3.2232, | |
| "step": 82200 | |
| }, | |
| { | |
| "epoch": 23.93822775966465, | |
| "grad_norm": 0.46508926153182983, | |
| "learning_rate": 0.0003129330227140361, | |
| "loss": 3.2214, | |
| "step": 82250 | |
| }, | |
| { | |
| "epoch": 23.952782952957616, | |
| "grad_norm": 0.4180763065814972, | |
| "learning_rate": 0.00031275829935934764, | |
| "loss": 3.2215, | |
| "step": 82300 | |
| }, | |
| { | |
| "epoch": 23.967338146250583, | |
| "grad_norm": 0.43056684732437134, | |
| "learning_rate": 0.00031258357600465924, | |
| "loss": 3.2203, | |
| "step": 82350 | |
| }, | |
| { | |
| "epoch": 23.98189333954355, | |
| "grad_norm": 0.4081912636756897, | |
| "learning_rate": 0.00031240885264997083, | |
| "loss": 3.2369, | |
| "step": 82400 | |
| }, | |
| { | |
| "epoch": 23.996448532836517, | |
| "grad_norm": 0.3891538083553314, | |
| "learning_rate": 0.0003122341292952825, | |
| "loss": 3.2197, | |
| "step": 82450 | |
| }, | |
| { | |
| "epoch": 24.010770843036795, | |
| "grad_norm": 0.42958930134773254, | |
| "learning_rate": 0.000312059405940594, | |
| "loss": 3.1498, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 24.025326036329762, | |
| "grad_norm": 0.4145827293395996, | |
| "learning_rate": 0.0003118846825859056, | |
| "loss": 3.1274, | |
| "step": 82550 | |
| }, | |
| { | |
| "epoch": 24.03988122962273, | |
| "grad_norm": 0.4117175042629242, | |
| "learning_rate": 0.0003117099592312172, | |
| "loss": 3.1281, | |
| "step": 82600 | |
| }, | |
| { | |
| "epoch": 24.054436422915696, | |
| "grad_norm": 0.4058437645435333, | |
| "learning_rate": 0.0003115352358765288, | |
| "loss": 3.1278, | |
| "step": 82650 | |
| }, | |
| { | |
| "epoch": 24.068991616208663, | |
| "grad_norm": 0.4439738094806671, | |
| "learning_rate": 0.0003113605125218404, | |
| "loss": 3.1437, | |
| "step": 82700 | |
| }, | |
| { | |
| "epoch": 24.08354680950163, | |
| "grad_norm": 0.4439554214477539, | |
| "learning_rate": 0.000311185789167152, | |
| "loss": 3.1419, | |
| "step": 82750 | |
| }, | |
| { | |
| "epoch": 24.098102002794597, | |
| "grad_norm": 0.42919254302978516, | |
| "learning_rate": 0.0003110110658124636, | |
| "loss": 3.146, | |
| "step": 82800 | |
| }, | |
| { | |
| "epoch": 24.112657196087564, | |
| "grad_norm": 0.4345749020576477, | |
| "learning_rate": 0.0003108363424577752, | |
| "loss": 3.1503, | |
| "step": 82850 | |
| }, | |
| { | |
| "epoch": 24.12721238938053, | |
| "grad_norm": 0.419748455286026, | |
| "learning_rate": 0.0003106616191030867, | |
| "loss": 3.1431, | |
| "step": 82900 | |
| }, | |
| { | |
| "epoch": 24.1417675826735, | |
| "grad_norm": 0.4344918429851532, | |
| "learning_rate": 0.0003104868957483983, | |
| "loss": 3.1408, | |
| "step": 82950 | |
| }, | |
| { | |
| "epoch": 24.156322775966466, | |
| "grad_norm": 0.4116227328777313, | |
| "learning_rate": 0.00031031217239370996, | |
| "loss": 3.1537, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.156322775966466, | |
| "eval_accuracy": 0.37389807867536384, | |
| "eval_loss": 3.5518624782562256, | |
| "eval_runtime": 181.3785, | |
| "eval_samples_per_second": 91.803, | |
| "eval_steps_per_second": 5.739, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.170877969259433, | |
| "grad_norm": 0.43057700991630554, | |
| "learning_rate": 0.00031013744903902156, | |
| "loss": 3.1558, | |
| "step": 83050 | |
| }, | |
| { | |
| "epoch": 24.1854331625524, | |
| "grad_norm": 0.4120100438594818, | |
| "learning_rate": 0.0003099627256843331, | |
| "loss": 3.1631, | |
| "step": 83100 | |
| }, | |
| { | |
| "epoch": 24.199988355845367, | |
| "grad_norm": 0.43784815073013306, | |
| "learning_rate": 0.0003097880023296447, | |
| "loss": 3.1501, | |
| "step": 83150 | |
| }, | |
| { | |
| "epoch": 24.214543549138334, | |
| "grad_norm": 0.41029947996139526, | |
| "learning_rate": 0.0003096132789749563, | |
| "loss": 3.1523, | |
| "step": 83200 | |
| }, | |
| { | |
| "epoch": 24.2290987424313, | |
| "grad_norm": 0.40311041474342346, | |
| "learning_rate": 0.00030943855562026794, | |
| "loss": 3.1533, | |
| "step": 83250 | |
| }, | |
| { | |
| "epoch": 24.243653935724268, | |
| "grad_norm": 0.42732328176498413, | |
| "learning_rate": 0.0003092638322655795, | |
| "loss": 3.1647, | |
| "step": 83300 | |
| }, | |
| { | |
| "epoch": 24.258209129017235, | |
| "grad_norm": 0.42801928520202637, | |
| "learning_rate": 0.00030908910891089107, | |
| "loss": 3.1621, | |
| "step": 83350 | |
| }, | |
| { | |
| "epoch": 24.2727643223102, | |
| "grad_norm": 0.4290964603424072, | |
| "learning_rate": 0.00030891438555620266, | |
| "loss": 3.1618, | |
| "step": 83400 | |
| }, | |
| { | |
| "epoch": 24.287319515603166, | |
| "grad_norm": 0.4055120348930359, | |
| "learning_rate": 0.0003087396622015142, | |
| "loss": 3.1756, | |
| "step": 83450 | |
| }, | |
| { | |
| "epoch": 24.301874708896133, | |
| "grad_norm": 0.42242538928985596, | |
| "learning_rate": 0.0003085649388468258, | |
| "loss": 3.1729, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 24.3164299021891, | |
| "grad_norm": 0.4095151126384735, | |
| "learning_rate": 0.00030839021549213745, | |
| "loss": 3.1704, | |
| "step": 83550 | |
| }, | |
| { | |
| "epoch": 24.330985095482067, | |
| "grad_norm": 0.40566420555114746, | |
| "learning_rate": 0.00030821549213744904, | |
| "loss": 3.166, | |
| "step": 83600 | |
| }, | |
| { | |
| "epoch": 24.345540288775034, | |
| "grad_norm": 0.3966980278491974, | |
| "learning_rate": 0.0003080407687827606, | |
| "loss": 3.1797, | |
| "step": 83650 | |
| }, | |
| { | |
| "epoch": 24.360095482068, | |
| "grad_norm": 0.4019142687320709, | |
| "learning_rate": 0.0003078660454280722, | |
| "loss": 3.1798, | |
| "step": 83700 | |
| }, | |
| { | |
| "epoch": 24.374650675360968, | |
| "grad_norm": 0.4082522988319397, | |
| "learning_rate": 0.00030769132207338377, | |
| "loss": 3.1807, | |
| "step": 83750 | |
| }, | |
| { | |
| "epoch": 24.389205868653935, | |
| "grad_norm": 0.4320712983608246, | |
| "learning_rate": 0.00030751659871869536, | |
| "loss": 3.1908, | |
| "step": 83800 | |
| }, | |
| { | |
| "epoch": 24.403761061946902, | |
| "grad_norm": 0.41094931960105896, | |
| "learning_rate": 0.00030734187536400696, | |
| "loss": 3.1735, | |
| "step": 83850 | |
| }, | |
| { | |
| "epoch": 24.41831625523987, | |
| "grad_norm": 0.42182329297065735, | |
| "learning_rate": 0.00030716715200931855, | |
| "loss": 3.1729, | |
| "step": 83900 | |
| }, | |
| { | |
| "epoch": 24.432871448532836, | |
| "grad_norm": 0.4036419689655304, | |
| "learning_rate": 0.00030699242865463015, | |
| "loss": 3.1887, | |
| "step": 83950 | |
| }, | |
| { | |
| "epoch": 24.447426641825803, | |
| "grad_norm": 0.41440510749816895, | |
| "learning_rate": 0.00030681770529994174, | |
| "loss": 3.1843, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.447426641825803, | |
| "eval_accuracy": 0.37409999058606286, | |
| "eval_loss": 3.544252395629883, | |
| "eval_runtime": 181.2073, | |
| "eval_samples_per_second": 91.889, | |
| "eval_steps_per_second": 5.745, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.46198183511877, | |
| "grad_norm": 0.4058787524700165, | |
| "learning_rate": 0.0003066429819452533, | |
| "loss": 3.1824, | |
| "step": 84050 | |
| }, | |
| { | |
| "epoch": 24.476537028411737, | |
| "grad_norm": 0.43140995502471924, | |
| "learning_rate": 0.00030646825859056493, | |
| "loss": 3.1743, | |
| "step": 84100 | |
| }, | |
| { | |
| "epoch": 24.491092221704704, | |
| "grad_norm": 0.4342149794101715, | |
| "learning_rate": 0.0003062935352358765, | |
| "loss": 3.1891, | |
| "step": 84150 | |
| }, | |
| { | |
| "epoch": 24.50564741499767, | |
| "grad_norm": 0.43598994612693787, | |
| "learning_rate": 0.0003061188118811881, | |
| "loss": 3.1846, | |
| "step": 84200 | |
| }, | |
| { | |
| "epoch": 24.52020260829064, | |
| "grad_norm": 0.41045066714286804, | |
| "learning_rate": 0.00030594408852649966, | |
| "loss": 3.1906, | |
| "step": 84250 | |
| }, | |
| { | |
| "epoch": 24.534757801583606, | |
| "grad_norm": 0.42254623770713806, | |
| "learning_rate": 0.00030576936517181125, | |
| "loss": 3.1885, | |
| "step": 84300 | |
| }, | |
| { | |
| "epoch": 24.549312994876573, | |
| "grad_norm": 0.4160739481449127, | |
| "learning_rate": 0.00030559464181712285, | |
| "loss": 3.1943, | |
| "step": 84350 | |
| }, | |
| { | |
| "epoch": 24.56386818816954, | |
| "grad_norm": 0.4158286452293396, | |
| "learning_rate": 0.0003054199184624345, | |
| "loss": 3.1982, | |
| "step": 84400 | |
| }, | |
| { | |
| "epoch": 24.578423381462507, | |
| "grad_norm": 0.4078032672405243, | |
| "learning_rate": 0.00030524519510774604, | |
| "loss": 3.1821, | |
| "step": 84450 | |
| }, | |
| { | |
| "epoch": 24.592978574755474, | |
| "grad_norm": 0.4104495644569397, | |
| "learning_rate": 0.00030507047175305763, | |
| "loss": 3.1941, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 24.60753376804844, | |
| "grad_norm": 0.42513424158096313, | |
| "learning_rate": 0.0003048957483983692, | |
| "loss": 3.1918, | |
| "step": 84550 | |
| }, | |
| { | |
| "epoch": 24.622088961341408, | |
| "grad_norm": 0.4459233582019806, | |
| "learning_rate": 0.00030472102504368076, | |
| "loss": 3.2031, | |
| "step": 84600 | |
| }, | |
| { | |
| "epoch": 24.636644154634375, | |
| "grad_norm": 0.39689961075782776, | |
| "learning_rate": 0.0003045463016889924, | |
| "loss": 3.1937, | |
| "step": 84650 | |
| }, | |
| { | |
| "epoch": 24.651199347927342, | |
| "grad_norm": 0.43621814250946045, | |
| "learning_rate": 0.000304371578334304, | |
| "loss": 3.201, | |
| "step": 84700 | |
| }, | |
| { | |
| "epoch": 24.66575454122031, | |
| "grad_norm": 0.3944084942340851, | |
| "learning_rate": 0.0003041968549796156, | |
| "loss": 3.1955, | |
| "step": 84750 | |
| }, | |
| { | |
| "epoch": 24.680309734513273, | |
| "grad_norm": 0.398531436920166, | |
| "learning_rate": 0.00030402213162492714, | |
| "loss": 3.1965, | |
| "step": 84800 | |
| }, | |
| { | |
| "epoch": 24.69486492780624, | |
| "grad_norm": 0.4374167323112488, | |
| "learning_rate": 0.00030384740827023874, | |
| "loss": 3.2008, | |
| "step": 84850 | |
| }, | |
| { | |
| "epoch": 24.709420121099207, | |
| "grad_norm": 0.41527554392814636, | |
| "learning_rate": 0.00030367268491555033, | |
| "loss": 3.185, | |
| "step": 84900 | |
| }, | |
| { | |
| "epoch": 24.723975314392174, | |
| "grad_norm": 0.4076023995876312, | |
| "learning_rate": 0.000303497961560862, | |
| "loss": 3.1992, | |
| "step": 84950 | |
| }, | |
| { | |
| "epoch": 24.73853050768514, | |
| "grad_norm": 0.4005376696586609, | |
| "learning_rate": 0.00030332323820617357, | |
| "loss": 3.2124, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.73853050768514, | |
| "eval_accuracy": 0.374507575281234, | |
| "eval_loss": 3.53737735748291, | |
| "eval_runtime": 181.2714, | |
| "eval_samples_per_second": 91.857, | |
| "eval_steps_per_second": 5.743, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.753085700978108, | |
| "grad_norm": 0.4087483286857605, | |
| "learning_rate": 0.0003031485148514851, | |
| "loss": 3.2045, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 24.767640894271075, | |
| "grad_norm": 0.4213216006755829, | |
| "learning_rate": 0.0003029737914967967, | |
| "loss": 3.2155, | |
| "step": 85100 | |
| }, | |
| { | |
| "epoch": 24.782196087564042, | |
| "grad_norm": 0.4251973330974579, | |
| "learning_rate": 0.0003027990681421083, | |
| "loss": 3.2027, | |
| "step": 85150 | |
| }, | |
| { | |
| "epoch": 24.79675128085701, | |
| "grad_norm": 0.4469214975833893, | |
| "learning_rate": 0.00030262434478741984, | |
| "loss": 3.2082, | |
| "step": 85200 | |
| }, | |
| { | |
| "epoch": 24.811306474149976, | |
| "grad_norm": 0.40554749965667725, | |
| "learning_rate": 0.0003024496214327315, | |
| "loss": 3.2054, | |
| "step": 85250 | |
| }, | |
| { | |
| "epoch": 24.825861667442943, | |
| "grad_norm": 0.4176938533782959, | |
| "learning_rate": 0.0003022748980780431, | |
| "loss": 3.1834, | |
| "step": 85300 | |
| }, | |
| { | |
| "epoch": 24.84041686073591, | |
| "grad_norm": 0.41417908668518066, | |
| "learning_rate": 0.0003021001747233547, | |
| "loss": 3.2208, | |
| "step": 85350 | |
| }, | |
| { | |
| "epoch": 24.854972054028877, | |
| "grad_norm": 0.42176830768585205, | |
| "learning_rate": 0.0003019254513686662, | |
| "loss": 3.2061, | |
| "step": 85400 | |
| }, | |
| { | |
| "epoch": 24.869527247321844, | |
| "grad_norm": 0.42138761281967163, | |
| "learning_rate": 0.0003017507280139778, | |
| "loss": 3.218, | |
| "step": 85450 | |
| }, | |
| { | |
| "epoch": 24.88408244061481, | |
| "grad_norm": 0.4482320249080658, | |
| "learning_rate": 0.00030157600465928946, | |
| "loss": 3.2105, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 24.89863763390778, | |
| "grad_norm": 0.42444801330566406, | |
| "learning_rate": 0.00030140128130460106, | |
| "loss": 3.218, | |
| "step": 85550 | |
| }, | |
| { | |
| "epoch": 24.913192827200746, | |
| "grad_norm": 0.39515721797943115, | |
| "learning_rate": 0.0003012265579499126, | |
| "loss": 3.2095, | |
| "step": 85600 | |
| }, | |
| { | |
| "epoch": 24.927748020493713, | |
| "grad_norm": 0.44244998693466187, | |
| "learning_rate": 0.0003010518345952242, | |
| "loss": 3.213, | |
| "step": 85650 | |
| }, | |
| { | |
| "epoch": 24.94230321378668, | |
| "grad_norm": 0.4148036539554596, | |
| "learning_rate": 0.0003008771112405358, | |
| "loss": 3.2165, | |
| "step": 85700 | |
| }, | |
| { | |
| "epoch": 24.956858407079647, | |
| "grad_norm": 0.4045441746711731, | |
| "learning_rate": 0.0003007023878858473, | |
| "loss": 3.2167, | |
| "step": 85750 | |
| }, | |
| { | |
| "epoch": 24.971413600372614, | |
| "grad_norm": 0.42188113927841187, | |
| "learning_rate": 0.000300527664531159, | |
| "loss": 3.2062, | |
| "step": 85800 | |
| }, | |
| { | |
| "epoch": 24.98596879366558, | |
| "grad_norm": 0.4339165985584259, | |
| "learning_rate": 0.00030035294117647057, | |
| "loss": 3.2123, | |
| "step": 85850 | |
| }, | |
| { | |
| "epoch": 25.00029110386586, | |
| "grad_norm": 0.42109426856040955, | |
| "learning_rate": 0.00030017821782178216, | |
| "loss": 3.2185, | |
| "step": 85900 | |
| }, | |
| { | |
| "epoch": 25.014846297158826, | |
| "grad_norm": 0.39719924330711365, | |
| "learning_rate": 0.00030000349446709376, | |
| "loss": 3.1139, | |
| "step": 85950 | |
| }, | |
| { | |
| "epoch": 25.029401490451793, | |
| "grad_norm": 0.39850518107414246, | |
| "learning_rate": 0.00029982877111240535, | |
| "loss": 3.1189, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.029401490451793, | |
| "eval_accuracy": 0.37425959266681325, | |
| "eval_loss": 3.546060800552368, | |
| "eval_runtime": 181.2175, | |
| "eval_samples_per_second": 91.884, | |
| "eval_steps_per_second": 5.744, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.04395668374476, | |
| "grad_norm": 0.43182504177093506, | |
| "learning_rate": 0.00029965404775771694, | |
| "loss": 3.1191, | |
| "step": 86050 | |
| }, | |
| { | |
| "epoch": 25.058511877037727, | |
| "grad_norm": 0.4257315695285797, | |
| "learning_rate": 0.0002994793244030285, | |
| "loss": 3.1306, | |
| "step": 86100 | |
| }, | |
| { | |
| "epoch": 25.073067070330694, | |
| "grad_norm": 0.426359623670578, | |
| "learning_rate": 0.00029930460104834013, | |
| "loss": 3.1286, | |
| "step": 86150 | |
| }, | |
| { | |
| "epoch": 25.08762226362366, | |
| "grad_norm": 0.4430175721645355, | |
| "learning_rate": 0.0002991298776936517, | |
| "loss": 3.1293, | |
| "step": 86200 | |
| }, | |
| { | |
| "epoch": 25.10217745691663, | |
| "grad_norm": 0.4329589605331421, | |
| "learning_rate": 0.00029895515433896327, | |
| "loss": 3.1417, | |
| "step": 86250 | |
| }, | |
| { | |
| "epoch": 25.116732650209595, | |
| "grad_norm": 0.4352037012577057, | |
| "learning_rate": 0.00029878043098427486, | |
| "loss": 3.1352, | |
| "step": 86300 | |
| }, | |
| { | |
| "epoch": 25.131287843502562, | |
| "grad_norm": 0.45002666115760803, | |
| "learning_rate": 0.00029860570762958646, | |
| "loss": 3.1452, | |
| "step": 86350 | |
| }, | |
| { | |
| "epoch": 25.14584303679553, | |
| "grad_norm": 0.4303296208381653, | |
| "learning_rate": 0.00029843098427489805, | |
| "loss": 3.1434, | |
| "step": 86400 | |
| }, | |
| { | |
| "epoch": 25.160398230088497, | |
| "grad_norm": 0.42453959584236145, | |
| "learning_rate": 0.00029825626092020964, | |
| "loss": 3.1406, | |
| "step": 86450 | |
| }, | |
| { | |
| "epoch": 25.174953423381464, | |
| "grad_norm": 0.4066697955131531, | |
| "learning_rate": 0.00029808153756552124, | |
| "loss": 3.1375, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 25.18950861667443, | |
| "grad_norm": 0.41828471422195435, | |
| "learning_rate": 0.00029790681421083283, | |
| "loss": 3.1537, | |
| "step": 86550 | |
| }, | |
| { | |
| "epoch": 25.204063809967398, | |
| "grad_norm": 0.4148005545139313, | |
| "learning_rate": 0.00029773209085614443, | |
| "loss": 3.1439, | |
| "step": 86600 | |
| }, | |
| { | |
| "epoch": 25.218619003260365, | |
| "grad_norm": 0.4200406074523926, | |
| "learning_rate": 0.00029755736750145597, | |
| "loss": 3.1465, | |
| "step": 86650 | |
| }, | |
| { | |
| "epoch": 25.233174196553332, | |
| "grad_norm": 0.41274210810661316, | |
| "learning_rate": 0.0002973826441467676, | |
| "loss": 3.1425, | |
| "step": 86700 | |
| }, | |
| { | |
| "epoch": 25.2477293898463, | |
| "grad_norm": 0.3914394974708557, | |
| "learning_rate": 0.00029720792079207916, | |
| "loss": 3.1553, | |
| "step": 86750 | |
| }, | |
| { | |
| "epoch": 25.262284583139262, | |
| "grad_norm": 0.41851523518562317, | |
| "learning_rate": 0.00029703319743739075, | |
| "loss": 3.1555, | |
| "step": 86800 | |
| }, | |
| { | |
| "epoch": 25.27683977643223, | |
| "grad_norm": 0.44641366600990295, | |
| "learning_rate": 0.00029685847408270234, | |
| "loss": 3.1664, | |
| "step": 86850 | |
| }, | |
| { | |
| "epoch": 25.291394969725197, | |
| "grad_norm": 0.4330465495586395, | |
| "learning_rate": 0.00029668375072801394, | |
| "loss": 3.1587, | |
| "step": 86900 | |
| }, | |
| { | |
| "epoch": 25.305950163018164, | |
| "grad_norm": 0.4315741956233978, | |
| "learning_rate": 0.00029650902737332553, | |
| "loss": 3.1571, | |
| "step": 86950 | |
| }, | |
| { | |
| "epoch": 25.32050535631113, | |
| "grad_norm": 0.41501671075820923, | |
| "learning_rate": 0.00029633430401863713, | |
| "loss": 3.1644, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.32050535631113, | |
| "eval_accuracy": 0.3738778639788329, | |
| "eval_loss": 3.550408124923706, | |
| "eval_runtime": 181.0862, | |
| "eval_samples_per_second": 91.951, | |
| "eval_steps_per_second": 5.749, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.335060549604098, | |
| "grad_norm": 0.430515319108963, | |
| "learning_rate": 0.0002961595806639487, | |
| "loss": 3.1494, | |
| "step": 87050 | |
| }, | |
| { | |
| "epoch": 25.349615742897065, | |
| "grad_norm": 0.4376870095729828, | |
| "learning_rate": 0.0002959848573092603, | |
| "loss": 3.1678, | |
| "step": 87100 | |
| }, | |
| { | |
| "epoch": 25.364170936190032, | |
| "grad_norm": 0.39572909474372864, | |
| "learning_rate": 0.0002958101339545719, | |
| "loss": 3.1846, | |
| "step": 87150 | |
| }, | |
| { | |
| "epoch": 25.378726129483, | |
| "grad_norm": 0.4197179079055786, | |
| "learning_rate": 0.0002956354105998835, | |
| "loss": 3.171, | |
| "step": 87200 | |
| }, | |
| { | |
| "epoch": 25.393281322775966, | |
| "grad_norm": 0.4425291419029236, | |
| "learning_rate": 0.0002954606872451951, | |
| "loss": 3.1659, | |
| "step": 87250 | |
| }, | |
| { | |
| "epoch": 25.407836516068933, | |
| "grad_norm": 0.4734668433666229, | |
| "learning_rate": 0.0002952859638905067, | |
| "loss": 3.1681, | |
| "step": 87300 | |
| }, | |
| { | |
| "epoch": 25.4223917093619, | |
| "grad_norm": 0.4288901090621948, | |
| "learning_rate": 0.00029511124053581823, | |
| "loss": 3.1686, | |
| "step": 87350 | |
| }, | |
| { | |
| "epoch": 25.436946902654867, | |
| "grad_norm": 0.41625797748565674, | |
| "learning_rate": 0.0002949365171811299, | |
| "loss": 3.1714, | |
| "step": 87400 | |
| }, | |
| { | |
| "epoch": 25.451502095947834, | |
| "grad_norm": 0.47367292642593384, | |
| "learning_rate": 0.0002947617938264414, | |
| "loss": 3.1676, | |
| "step": 87450 | |
| }, | |
| { | |
| "epoch": 25.4660572892408, | |
| "grad_norm": 0.42922917008399963, | |
| "learning_rate": 0.000294587070471753, | |
| "loss": 3.1726, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 25.48061248253377, | |
| "grad_norm": 0.42955517768859863, | |
| "learning_rate": 0.0002944123471170646, | |
| "loss": 3.186, | |
| "step": 87550 | |
| }, | |
| { | |
| "epoch": 25.495167675826735, | |
| "grad_norm": 0.4782566726207733, | |
| "learning_rate": 0.0002942376237623762, | |
| "loss": 3.197, | |
| "step": 87600 | |
| }, | |
| { | |
| "epoch": 25.509722869119702, | |
| "grad_norm": 0.43603309988975525, | |
| "learning_rate": 0.0002940629004076878, | |
| "loss": 3.1809, | |
| "step": 87650 | |
| }, | |
| { | |
| "epoch": 25.52427806241267, | |
| "grad_norm": 0.44062429666519165, | |
| "learning_rate": 0.0002938881770529994, | |
| "loss": 3.1796, | |
| "step": 87700 | |
| }, | |
| { | |
| "epoch": 25.538833255705637, | |
| "grad_norm": 0.39834269881248474, | |
| "learning_rate": 0.000293713453698311, | |
| "loss": 3.1853, | |
| "step": 87750 | |
| }, | |
| { | |
| "epoch": 25.553388448998604, | |
| "grad_norm": 0.4281576871871948, | |
| "learning_rate": 0.0002935387303436226, | |
| "loss": 3.1779, | |
| "step": 87800 | |
| }, | |
| { | |
| "epoch": 25.56794364229157, | |
| "grad_norm": 0.4472132623195648, | |
| "learning_rate": 0.0002933640069889342, | |
| "loss": 3.1837, | |
| "step": 87850 | |
| }, | |
| { | |
| "epoch": 25.582498835584538, | |
| "grad_norm": 0.42856264114379883, | |
| "learning_rate": 0.0002931892836342457, | |
| "loss": 3.1878, | |
| "step": 87900 | |
| }, | |
| { | |
| "epoch": 25.597054028877505, | |
| "grad_norm": 0.4264664649963379, | |
| "learning_rate": 0.00029301456027955736, | |
| "loss": 3.1873, | |
| "step": 87950 | |
| }, | |
| { | |
| "epoch": 25.611609222170472, | |
| "grad_norm": 0.40437066555023193, | |
| "learning_rate": 0.0002928398369248689, | |
| "loss": 3.1798, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.611609222170472, | |
| "eval_accuracy": 0.374592429995742, | |
| "eval_loss": 3.5399508476257324, | |
| "eval_runtime": 181.2957, | |
| "eval_samples_per_second": 91.844, | |
| "eval_steps_per_second": 5.742, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.62616441546344, | |
| "grad_norm": 0.466431587934494, | |
| "learning_rate": 0.0002926651135701805, | |
| "loss": 3.1867, | |
| "step": 88050 | |
| }, | |
| { | |
| "epoch": 25.640719608756406, | |
| "grad_norm": 0.43069714307785034, | |
| "learning_rate": 0.00029249039021549215, | |
| "loss": 3.1875, | |
| "step": 88100 | |
| }, | |
| { | |
| "epoch": 25.65527480204937, | |
| "grad_norm": 0.4674331843852997, | |
| "learning_rate": 0.0002923156668608037, | |
| "loss": 3.1889, | |
| "step": 88150 | |
| }, | |
| { | |
| "epoch": 25.669829995342337, | |
| "grad_norm": 0.42163434624671936, | |
| "learning_rate": 0.0002921409435061153, | |
| "loss": 3.1871, | |
| "step": 88200 | |
| }, | |
| { | |
| "epoch": 25.684385188635304, | |
| "grad_norm": 0.4694327116012573, | |
| "learning_rate": 0.0002919662201514269, | |
| "loss": 3.1816, | |
| "step": 88250 | |
| }, | |
| { | |
| "epoch": 25.69894038192827, | |
| "grad_norm": 0.43342265486717224, | |
| "learning_rate": 0.00029179149679673847, | |
| "loss": 3.183, | |
| "step": 88300 | |
| }, | |
| { | |
| "epoch": 25.713495575221238, | |
| "grad_norm": 0.436389684677124, | |
| "learning_rate": 0.00029161677344205007, | |
| "loss": 3.1894, | |
| "step": 88350 | |
| }, | |
| { | |
| "epoch": 25.728050768514205, | |
| "grad_norm": 0.4253075420856476, | |
| "learning_rate": 0.00029144205008736166, | |
| "loss": 3.1994, | |
| "step": 88400 | |
| }, | |
| { | |
| "epoch": 25.742605961807172, | |
| "grad_norm": 0.4468969404697418, | |
| "learning_rate": 0.00029126732673267325, | |
| "loss": 3.2013, | |
| "step": 88450 | |
| }, | |
| { | |
| "epoch": 25.75716115510014, | |
| "grad_norm": 0.3974677324295044, | |
| "learning_rate": 0.00029109260337798485, | |
| "loss": 3.1971, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 25.771716348393106, | |
| "grad_norm": 0.4659837782382965, | |
| "learning_rate": 0.00029091788002329644, | |
| "loss": 3.1899, | |
| "step": 88550 | |
| }, | |
| { | |
| "epoch": 25.786271541686073, | |
| "grad_norm": 0.423094779253006, | |
| "learning_rate": 0.000290743156668608, | |
| "loss": 3.1907, | |
| "step": 88600 | |
| }, | |
| { | |
| "epoch": 25.80082673497904, | |
| "grad_norm": 0.4330526292324066, | |
| "learning_rate": 0.00029056843331391963, | |
| "loss": 3.2041, | |
| "step": 88650 | |
| }, | |
| { | |
| "epoch": 25.815381928272007, | |
| "grad_norm": 0.4233783185482025, | |
| "learning_rate": 0.00029039370995923117, | |
| "loss": 3.1941, | |
| "step": 88700 | |
| }, | |
| { | |
| "epoch": 25.829937121564974, | |
| "grad_norm": 0.4054449498653412, | |
| "learning_rate": 0.00029021898660454277, | |
| "loss": 3.2012, | |
| "step": 88750 | |
| }, | |
| { | |
| "epoch": 25.84449231485794, | |
| "grad_norm": 0.4207738935947418, | |
| "learning_rate": 0.00029004426324985436, | |
| "loss": 3.1995, | |
| "step": 88800 | |
| }, | |
| { | |
| "epoch": 25.85904750815091, | |
| "grad_norm": 0.3913928270339966, | |
| "learning_rate": 0.00028986953989516595, | |
| "loss": 3.1966, | |
| "step": 88850 | |
| }, | |
| { | |
| "epoch": 25.873602701443875, | |
| "grad_norm": 0.41812190413475037, | |
| "learning_rate": 0.00028969481654047755, | |
| "loss": 3.2021, | |
| "step": 88900 | |
| }, | |
| { | |
| "epoch": 25.888157894736842, | |
| "grad_norm": 0.42711809277534485, | |
| "learning_rate": 0.00028952009318578914, | |
| "loss": 3.2104, | |
| "step": 88950 | |
| }, | |
| { | |
| "epoch": 25.90271308802981, | |
| "grad_norm": 0.41786685585975647, | |
| "learning_rate": 0.00028934536983110074, | |
| "loss": 3.2071, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.90271308802981, | |
| "eval_accuracy": 0.3751760705944214, | |
| "eval_loss": 3.532879114151001, | |
| "eval_runtime": 181.0754, | |
| "eval_samples_per_second": 91.956, | |
| "eval_steps_per_second": 5.749, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.917268281322777, | |
| "grad_norm": 0.41919800639152527, | |
| "learning_rate": 0.00028917064647641233, | |
| "loss": 3.2159, | |
| "step": 89050 | |
| }, | |
| { | |
| "epoch": 25.931823474615744, | |
| "grad_norm": 0.4346853792667389, | |
| "learning_rate": 0.0002889959231217239, | |
| "loss": 3.2086, | |
| "step": 89100 | |
| }, | |
| { | |
| "epoch": 25.94637866790871, | |
| "grad_norm": 0.42549893260002136, | |
| "learning_rate": 0.0002888211997670355, | |
| "loss": 3.1917, | |
| "step": 89150 | |
| }, | |
| { | |
| "epoch": 25.960933861201678, | |
| "grad_norm": 0.42451906204223633, | |
| "learning_rate": 0.0002886464764123471, | |
| "loss": 3.2051, | |
| "step": 89200 | |
| }, | |
| { | |
| "epoch": 25.975489054494645, | |
| "grad_norm": 0.4404706358909607, | |
| "learning_rate": 0.0002884717530576587, | |
| "loss": 3.203, | |
| "step": 89250 | |
| }, | |
| { | |
| "epoch": 25.990044247787612, | |
| "grad_norm": 0.41028639674186707, | |
| "learning_rate": 0.00028829702970297025, | |
| "loss": 3.1926, | |
| "step": 89300 | |
| }, | |
| { | |
| "epoch": 26.00436655798789, | |
| "grad_norm": 0.48599815368652344, | |
| "learning_rate": 0.0002881223063482819, | |
| "loss": 3.1672, | |
| "step": 89350 | |
| }, | |
| { | |
| "epoch": 26.018921751280857, | |
| "grad_norm": 0.42923834919929504, | |
| "learning_rate": 0.00028794758299359344, | |
| "loss": 3.102, | |
| "step": 89400 | |
| }, | |
| { | |
| "epoch": 26.033476944573824, | |
| "grad_norm": 0.42139095067977905, | |
| "learning_rate": 0.00028777285963890503, | |
| "loss": 3.1074, | |
| "step": 89450 | |
| }, | |
| { | |
| "epoch": 26.04803213786679, | |
| "grad_norm": 0.42988842725753784, | |
| "learning_rate": 0.0002875981362842166, | |
| "loss": 3.1249, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 26.062587331159758, | |
| "grad_norm": 0.4336186945438385, | |
| "learning_rate": 0.0002874234129295282, | |
| "loss": 3.1217, | |
| "step": 89550 | |
| }, | |
| { | |
| "epoch": 26.077142524452725, | |
| "grad_norm": 0.40431252121925354, | |
| "learning_rate": 0.0002872486895748398, | |
| "loss": 3.13, | |
| "step": 89600 | |
| }, | |
| { | |
| "epoch": 26.091697717745692, | |
| "grad_norm": 0.4139500856399536, | |
| "learning_rate": 0.0002870739662201514, | |
| "loss": 3.13, | |
| "step": 89650 | |
| }, | |
| { | |
| "epoch": 26.10625291103866, | |
| "grad_norm": 0.4227358400821686, | |
| "learning_rate": 0.000286899242865463, | |
| "loss": 3.119, | |
| "step": 89700 | |
| }, | |
| { | |
| "epoch": 26.120808104331626, | |
| "grad_norm": 0.41390228271484375, | |
| "learning_rate": 0.0002867245195107746, | |
| "loss": 3.1207, | |
| "step": 89750 | |
| }, | |
| { | |
| "epoch": 26.135363297624593, | |
| "grad_norm": 0.4205567538738251, | |
| "learning_rate": 0.0002865497961560862, | |
| "loss": 3.1274, | |
| "step": 89800 | |
| }, | |
| { | |
| "epoch": 26.14991849091756, | |
| "grad_norm": 0.43315389752388, | |
| "learning_rate": 0.00028637507280139773, | |
| "loss": 3.1347, | |
| "step": 89850 | |
| }, | |
| { | |
| "epoch": 26.164473684210527, | |
| "grad_norm": 0.4300045967102051, | |
| "learning_rate": 0.0002862003494467094, | |
| "loss": 3.1347, | |
| "step": 89900 | |
| }, | |
| { | |
| "epoch": 26.179028877503494, | |
| "grad_norm": 0.419502854347229, | |
| "learning_rate": 0.0002860256260920209, | |
| "loss": 3.14, | |
| "step": 89950 | |
| }, | |
| { | |
| "epoch": 26.19358407079646, | |
| "grad_norm": 0.4483463764190674, | |
| "learning_rate": 0.0002858509027373325, | |
| "loss": 3.1338, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.19358407079646, | |
| "eval_accuracy": 0.373934864721958, | |
| "eval_loss": 3.5498576164245605, | |
| "eval_runtime": 180.9451, | |
| "eval_samples_per_second": 92.022, | |
| "eval_steps_per_second": 5.753, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.207848160223566, | |
| "grad_norm": 0.4221210777759552, | |
| "learning_rate": 0.0002856761793826441, | |
| "loss": 3.1261, | |
| "step": 90050 | |
| }, | |
| { | |
| "epoch": 26.222403353516533, | |
| "grad_norm": 0.42945337295532227, | |
| "learning_rate": 0.0002855014560279557, | |
| "loss": 3.1235, | |
| "step": 90100 | |
| }, | |
| { | |
| "epoch": 26.2369585468095, | |
| "grad_norm": 0.421125590801239, | |
| "learning_rate": 0.0002853267326732673, | |
| "loss": 3.1238, | |
| "step": 90150 | |
| }, | |
| { | |
| "epoch": 26.251513740102467, | |
| "grad_norm": 0.43803295493125916, | |
| "learning_rate": 0.0002851520093185789, | |
| "loss": 3.1205, | |
| "step": 90200 | |
| }, | |
| { | |
| "epoch": 26.266068933395434, | |
| "grad_norm": 0.43245306611061096, | |
| "learning_rate": 0.0002849772859638905, | |
| "loss": 3.1183, | |
| "step": 90250 | |
| }, | |
| { | |
| "epoch": 26.2806241266884, | |
| "grad_norm": 0.43539243936538696, | |
| "learning_rate": 0.0002848025626092021, | |
| "loss": 3.1116, | |
| "step": 90300 | |
| }, | |
| { | |
| "epoch": 26.29517931998137, | |
| "grad_norm": 0.45199355483055115, | |
| "learning_rate": 0.0002846278392545137, | |
| "loss": 3.1217, | |
| "step": 90350 | |
| }, | |
| { | |
| "epoch": 26.309734513274336, | |
| "grad_norm": 0.44650569558143616, | |
| "learning_rate": 0.00028445311589982527, | |
| "loss": 3.1276, | |
| "step": 90400 | |
| }, | |
| { | |
| "epoch": 26.324289706567303, | |
| "grad_norm": 0.4259169101715088, | |
| "learning_rate": 0.00028427839254513686, | |
| "loss": 3.1345, | |
| "step": 90450 | |
| }, | |
| { | |
| "epoch": 26.33884489986027, | |
| "grad_norm": 0.4632343053817749, | |
| "learning_rate": 0.00028410366919044846, | |
| "loss": 3.132, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 26.353400093153237, | |
| "grad_norm": 0.43775540590286255, | |
| "learning_rate": 0.00028392894583576, | |
| "loss": 3.1356, | |
| "step": 90550 | |
| }, | |
| { | |
| "epoch": 26.367955286446204, | |
| "grad_norm": 0.431986927986145, | |
| "learning_rate": 0.00028375422248107165, | |
| "loss": 3.1475, | |
| "step": 90600 | |
| }, | |
| { | |
| "epoch": 26.38251047973917, | |
| "grad_norm": 0.4202793538570404, | |
| "learning_rate": 0.0002835794991263832, | |
| "loss": 3.1447, | |
| "step": 90650 | |
| }, | |
| { | |
| "epoch": 26.397065673032138, | |
| "grad_norm": 0.4278116524219513, | |
| "learning_rate": 0.0002834047757716948, | |
| "loss": 3.1459, | |
| "step": 90700 | |
| }, | |
| { | |
| "epoch": 26.411620866325105, | |
| "grad_norm": 0.4313576817512512, | |
| "learning_rate": 0.0002832300524170064, | |
| "loss": 3.1412, | |
| "step": 90750 | |
| }, | |
| { | |
| "epoch": 26.426176059618072, | |
| "grad_norm": 0.45625731348991394, | |
| "learning_rate": 0.00028305532906231797, | |
| "loss": 3.1424, | |
| "step": 90800 | |
| }, | |
| { | |
| "epoch": 26.44073125291104, | |
| "grad_norm": 0.4196455776691437, | |
| "learning_rate": 0.00028288060570762956, | |
| "loss": 3.156, | |
| "step": 90850 | |
| }, | |
| { | |
| "epoch": 26.455286446204006, | |
| "grad_norm": 0.40867993235588074, | |
| "learning_rate": 0.00028270588235294116, | |
| "loss": 3.1573, | |
| "step": 90900 | |
| }, | |
| { | |
| "epoch": 26.469841639496973, | |
| "grad_norm": 0.4437214136123657, | |
| "learning_rate": 0.00028253115899825275, | |
| "loss": 3.1435, | |
| "step": 90950 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "grad_norm": 0.4219045341014862, | |
| "learning_rate": 0.0002823564356435643, | |
| "loss": 3.1636, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "eval_accuracy": 0.3740536848277302, | |
| "eval_loss": 3.549358606338501, | |
| "eval_runtime": 182.1405, | |
| "eval_samples_per_second": 91.418, | |
| "eval_steps_per_second": 5.715, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.498952026082907, | |
| "grad_norm": 0.4574682116508484, | |
| "learning_rate": 0.00028218171228887594, | |
| "loss": 3.1436, | |
| "step": 91050 | |
| }, | |
| { | |
| "epoch": 26.513507219375875, | |
| "grad_norm": 0.41283056139945984, | |
| "learning_rate": 0.0002820069889341875, | |
| "loss": 3.1572, | |
| "step": 91100 | |
| }, | |
| { | |
| "epoch": 26.52806241266884, | |
| "grad_norm": 0.4420188069343567, | |
| "learning_rate": 0.00028183226557949913, | |
| "loss": 3.1582, | |
| "step": 91150 | |
| }, | |
| { | |
| "epoch": 26.54261760596181, | |
| "grad_norm": 0.42842113971710205, | |
| "learning_rate": 0.00028165754222481067, | |
| "loss": 3.1653, | |
| "step": 91200 | |
| }, | |
| { | |
| "epoch": 26.557172799254776, | |
| "grad_norm": 0.4411379396915436, | |
| "learning_rate": 0.00028148281887012226, | |
| "loss": 3.1639, | |
| "step": 91250 | |
| }, | |
| { | |
| "epoch": 26.571727992547743, | |
| "grad_norm": 0.4473015069961548, | |
| "learning_rate": 0.0002813080955154339, | |
| "loss": 3.1546, | |
| "step": 91300 | |
| }, | |
| { | |
| "epoch": 26.586283185840706, | |
| "grad_norm": 0.4217207729816437, | |
| "learning_rate": 0.00028113337216074545, | |
| "loss": 3.1581, | |
| "step": 91350 | |
| }, | |
| { | |
| "epoch": 26.600838379133673, | |
| "grad_norm": 0.43184933066368103, | |
| "learning_rate": 0.00028095864880605705, | |
| "loss": 3.153, | |
| "step": 91400 | |
| }, | |
| { | |
| "epoch": 26.61539357242664, | |
| "grad_norm": 0.4376067817211151, | |
| "learning_rate": 0.00028078392545136864, | |
| "loss": 3.1572, | |
| "step": 91450 | |
| }, | |
| { | |
| "epoch": 26.629948765719607, | |
| "grad_norm": 0.46065959334373474, | |
| "learning_rate": 0.00028060920209668023, | |
| "loss": 3.1616, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 26.644503959012575, | |
| "grad_norm": 0.42812418937683105, | |
| "learning_rate": 0.00028043447874199183, | |
| "loss": 3.1661, | |
| "step": 91550 | |
| }, | |
| { | |
| "epoch": 26.65905915230554, | |
| "grad_norm": 0.44456830620765686, | |
| "learning_rate": 0.0002802597553873034, | |
| "loss": 3.1581, | |
| "step": 91600 | |
| }, | |
| { | |
| "epoch": 26.67361434559851, | |
| "grad_norm": 0.4181954860687256, | |
| "learning_rate": 0.000280085032032615, | |
| "loss": 3.1771, | |
| "step": 91650 | |
| }, | |
| { | |
| "epoch": 26.688169538891476, | |
| "grad_norm": 0.4161057770252228, | |
| "learning_rate": 0.00027991030867792656, | |
| "loss": 3.1825, | |
| "step": 91700 | |
| }, | |
| { | |
| "epoch": 26.702724732184443, | |
| "grad_norm": 0.41539129614830017, | |
| "learning_rate": 0.0002797355853232382, | |
| "loss": 3.1715, | |
| "step": 91750 | |
| }, | |
| { | |
| "epoch": 26.71727992547741, | |
| "grad_norm": 0.43947649002075195, | |
| "learning_rate": 0.00027956086196854975, | |
| "loss": 3.1684, | |
| "step": 91800 | |
| }, | |
| { | |
| "epoch": 26.731835118770377, | |
| "grad_norm": 0.4386477768421173, | |
| "learning_rate": 0.0002793861386138614, | |
| "loss": 3.1752, | |
| "step": 91850 | |
| }, | |
| { | |
| "epoch": 26.746390312063344, | |
| "grad_norm": 0.46494022011756897, | |
| "learning_rate": 0.00027921141525917293, | |
| "loss": 3.1694, | |
| "step": 91900 | |
| }, | |
| { | |
| "epoch": 26.76094550535631, | |
| "grad_norm": 0.43608811497688293, | |
| "learning_rate": 0.00027903669190448453, | |
| "loss": 3.1695, | |
| "step": 91950 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "grad_norm": 0.43629440665245056, | |
| "learning_rate": 0.0002788619685497961, | |
| "loss": 3.1776, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "eval_accuracy": 0.37460982403694304, | |
| "eval_loss": 3.5446009635925293, | |
| "eval_runtime": 181.4344, | |
| "eval_samples_per_second": 91.774, | |
| "eval_steps_per_second": 5.738, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.790055891942245, | |
| "grad_norm": 0.4232460558414459, | |
| "learning_rate": 0.0002786872451951077, | |
| "loss": 3.1696, | |
| "step": 92050 | |
| }, | |
| { | |
| "epoch": 26.804611085235212, | |
| "grad_norm": 0.42919155955314636, | |
| "learning_rate": 0.0002785125218404193, | |
| "loss": 3.175, | |
| "step": 92100 | |
| }, | |
| { | |
| "epoch": 26.81916627852818, | |
| "grad_norm": 0.42091572284698486, | |
| "learning_rate": 0.0002783377984857309, | |
| "loss": 3.1751, | |
| "step": 92150 | |
| }, | |
| { | |
| "epoch": 26.833721471821146, | |
| "grad_norm": 0.4050874412059784, | |
| "learning_rate": 0.0002781630751310425, | |
| "loss": 3.1813, | |
| "step": 92200 | |
| }, | |
| { | |
| "epoch": 26.848276665114113, | |
| "grad_norm": 0.4387812614440918, | |
| "learning_rate": 0.0002779883517763541, | |
| "loss": 3.1837, | |
| "step": 92250 | |
| }, | |
| { | |
| "epoch": 26.86283185840708, | |
| "grad_norm": 0.4179288446903229, | |
| "learning_rate": 0.0002778136284216657, | |
| "loss": 3.181, | |
| "step": 92300 | |
| }, | |
| { | |
| "epoch": 26.877387051700047, | |
| "grad_norm": 0.42394018173217773, | |
| "learning_rate": 0.0002776389050669773, | |
| "loss": 3.1847, | |
| "step": 92350 | |
| }, | |
| { | |
| "epoch": 26.891942244993015, | |
| "grad_norm": 0.44571471214294434, | |
| "learning_rate": 0.0002774641817122888, | |
| "loss": 3.2013, | |
| "step": 92400 | |
| }, | |
| { | |
| "epoch": 26.90649743828598, | |
| "grad_norm": 0.4174729287624359, | |
| "learning_rate": 0.00027728945835760047, | |
| "loss": 3.187, | |
| "step": 92450 | |
| }, | |
| { | |
| "epoch": 26.92105263157895, | |
| "grad_norm": 0.45103052258491516, | |
| "learning_rate": 0.000277114735002912, | |
| "loss": 3.175, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 26.935607824871916, | |
| "grad_norm": 0.42905762791633606, | |
| "learning_rate": 0.00027694001164822366, | |
| "loss": 3.1818, | |
| "step": 92550 | |
| }, | |
| { | |
| "epoch": 26.950163018164883, | |
| "grad_norm": 0.45374882221221924, | |
| "learning_rate": 0.0002767652882935352, | |
| "loss": 3.1782, | |
| "step": 92600 | |
| }, | |
| { | |
| "epoch": 26.96471821145785, | |
| "grad_norm": 0.44403746724128723, | |
| "learning_rate": 0.0002765905649388468, | |
| "loss": 3.1673, | |
| "step": 92650 | |
| }, | |
| { | |
| "epoch": 26.979273404750813, | |
| "grad_norm": 0.41546887159347534, | |
| "learning_rate": 0.0002764158415841584, | |
| "loss": 3.1826, | |
| "step": 92700 | |
| }, | |
| { | |
| "epoch": 26.99382859804378, | |
| "grad_norm": 0.41803351044654846, | |
| "learning_rate": 0.00027624111822947, | |
| "loss": 3.1781, | |
| "step": 92750 | |
| }, | |
| { | |
| "epoch": 27.00844201210992, | |
| "grad_norm": 0.4286218285560608, | |
| "learning_rate": 0.0002760663948747816, | |
| "loss": 3.1922, | |
| "step": 92800 | |
| }, | |
| { | |
| "epoch": 27.022997205402888, | |
| "grad_norm": 0.4495951235294342, | |
| "learning_rate": 0.00027589167152009317, | |
| "loss": 3.1048, | |
| "step": 92850 | |
| }, | |
| { | |
| "epoch": 27.037552398695855, | |
| "grad_norm": 0.4294961392879486, | |
| "learning_rate": 0.00027571694816540477, | |
| "loss": 3.1089, | |
| "step": 92900 | |
| }, | |
| { | |
| "epoch": 27.052107591988822, | |
| "grad_norm": 0.44172975420951843, | |
| "learning_rate": 0.0002755422248107163, | |
| "loss": 3.1145, | |
| "step": 92950 | |
| }, | |
| { | |
| "epoch": 27.06666278528179, | |
| "grad_norm": 0.40919336676597595, | |
| "learning_rate": 0.00027536750145602795, | |
| "loss": 3.1153, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.06666278528179, | |
| "eval_accuracy": 0.3738698721220648, | |
| "eval_loss": 3.5517330169677734, | |
| "eval_runtime": 181.3112, | |
| "eval_samples_per_second": 91.837, | |
| "eval_steps_per_second": 5.742, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.081217978574756, | |
| "grad_norm": 0.4204716682434082, | |
| "learning_rate": 0.0002751927781013395, | |
| "loss": 3.1158, | |
| "step": 93050 | |
| }, | |
| { | |
| "epoch": 27.095773171867723, | |
| "grad_norm": 0.4470804035663605, | |
| "learning_rate": 0.0002750180547466511, | |
| "loss": 3.1204, | |
| "step": 93100 | |
| }, | |
| { | |
| "epoch": 27.11032836516069, | |
| "grad_norm": 0.4344484508037567, | |
| "learning_rate": 0.0002748433313919627, | |
| "loss": 3.1196, | |
| "step": 93150 | |
| }, | |
| { | |
| "epoch": 27.124883558453657, | |
| "grad_norm": 0.4729747474193573, | |
| "learning_rate": 0.0002746686080372743, | |
| "loss": 3.123, | |
| "step": 93200 | |
| }, | |
| { | |
| "epoch": 27.139438751746624, | |
| "grad_norm": 0.40821605920791626, | |
| "learning_rate": 0.00027449388468258587, | |
| "loss": 3.1364, | |
| "step": 93250 | |
| }, | |
| { | |
| "epoch": 27.15399394503959, | |
| "grad_norm": 0.4483500123023987, | |
| "learning_rate": 0.00027431916132789747, | |
| "loss": 3.1251, | |
| "step": 93300 | |
| }, | |
| { | |
| "epoch": 27.16854913833256, | |
| "grad_norm": 0.42222732305526733, | |
| "learning_rate": 0.00027414443797320906, | |
| "loss": 3.1326, | |
| "step": 93350 | |
| }, | |
| { | |
| "epoch": 27.183104331625525, | |
| "grad_norm": 0.41381341218948364, | |
| "learning_rate": 0.00027396971461852065, | |
| "loss": 3.1297, | |
| "step": 93400 | |
| }, | |
| { | |
| "epoch": 27.197659524918492, | |
| "grad_norm": 0.4309457838535309, | |
| "learning_rate": 0.00027379499126383225, | |
| "loss": 3.1373, | |
| "step": 93450 | |
| }, | |
| { | |
| "epoch": 27.21221471821146, | |
| "grad_norm": 0.42898011207580566, | |
| "learning_rate": 0.00027362026790914384, | |
| "loss": 3.1305, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 27.226769911504423, | |
| "grad_norm": 0.449894517660141, | |
| "learning_rate": 0.00027344554455445544, | |
| "loss": 3.1401, | |
| "step": 93550 | |
| }, | |
| { | |
| "epoch": 27.24132510479739, | |
| "grad_norm": 0.4240707457065582, | |
| "learning_rate": 0.00027327082119976703, | |
| "loss": 3.138, | |
| "step": 93600 | |
| }, | |
| { | |
| "epoch": 27.255880298090357, | |
| "grad_norm": 0.4510543942451477, | |
| "learning_rate": 0.00027309609784507857, | |
| "loss": 3.1454, | |
| "step": 93650 | |
| }, | |
| { | |
| "epoch": 27.270435491383324, | |
| "grad_norm": 0.41815170645713806, | |
| "learning_rate": 0.0002729213744903902, | |
| "loss": 3.1415, | |
| "step": 93700 | |
| }, | |
| { | |
| "epoch": 27.28499068467629, | |
| "grad_norm": 0.4698253273963928, | |
| "learning_rate": 0.00027274665113570176, | |
| "loss": 3.1491, | |
| "step": 93750 | |
| }, | |
| { | |
| "epoch": 27.29954587796926, | |
| "grad_norm": 0.44704052805900574, | |
| "learning_rate": 0.00027257192778101335, | |
| "loss": 3.1542, | |
| "step": 93800 | |
| }, | |
| { | |
| "epoch": 27.314101071262225, | |
| "grad_norm": 0.4419684112071991, | |
| "learning_rate": 0.00027239720442632495, | |
| "loss": 3.1361, | |
| "step": 93850 | |
| }, | |
| { | |
| "epoch": 27.328656264555192, | |
| "grad_norm": 0.4628223776817322, | |
| "learning_rate": 0.00027222248107163654, | |
| "loss": 3.1476, | |
| "step": 93900 | |
| }, | |
| { | |
| "epoch": 27.34321145784816, | |
| "grad_norm": 0.43916580080986023, | |
| "learning_rate": 0.00027204775771694814, | |
| "loss": 3.1373, | |
| "step": 93950 | |
| }, | |
| { | |
| "epoch": 27.357766651141127, | |
| "grad_norm": 0.40940552949905396, | |
| "learning_rate": 0.00027187303436225973, | |
| "loss": 3.1558, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.357766651141127, | |
| "eval_accuracy": 0.37423256138656835, | |
| "eval_loss": 3.548722743988037, | |
| "eval_runtime": 181.1383, | |
| "eval_samples_per_second": 91.924, | |
| "eval_steps_per_second": 5.747, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.372321844434094, | |
| "grad_norm": 0.4344089925289154, | |
| "learning_rate": 0.0002716983110075713, | |
| "loss": 3.1457, | |
| "step": 94050 | |
| }, | |
| { | |
| "epoch": 27.38687703772706, | |
| "grad_norm": 0.4405762553215027, | |
| "learning_rate": 0.0002715235876528829, | |
| "loss": 3.1544, | |
| "step": 94100 | |
| }, | |
| { | |
| "epoch": 27.401432231020028, | |
| "grad_norm": 0.4617936611175537, | |
| "learning_rate": 0.0002713488642981945, | |
| "loss": 3.1517, | |
| "step": 94150 | |
| }, | |
| { | |
| "epoch": 27.415987424312995, | |
| "grad_norm": 0.4110572934150696, | |
| "learning_rate": 0.00027117414094350606, | |
| "loss": 3.162, | |
| "step": 94200 | |
| }, | |
| { | |
| "epoch": 27.430542617605962, | |
| "grad_norm": 0.4259830713272095, | |
| "learning_rate": 0.0002709994175888177, | |
| "loss": 3.1644, | |
| "step": 94250 | |
| }, | |
| { | |
| "epoch": 27.44509781089893, | |
| "grad_norm": 0.46111369132995605, | |
| "learning_rate": 0.00027082469423412924, | |
| "loss": 3.1597, | |
| "step": 94300 | |
| }, | |
| { | |
| "epoch": 27.459653004191896, | |
| "grad_norm": 0.4082739055156708, | |
| "learning_rate": 0.00027064997087944084, | |
| "loss": 3.1719, | |
| "step": 94350 | |
| }, | |
| { | |
| "epoch": 27.474208197484863, | |
| "grad_norm": 0.40091511607170105, | |
| "learning_rate": 0.00027047524752475243, | |
| "loss": 3.1593, | |
| "step": 94400 | |
| }, | |
| { | |
| "epoch": 27.48876339077783, | |
| "grad_norm": 0.4257362484931946, | |
| "learning_rate": 0.000270300524170064, | |
| "loss": 3.1685, | |
| "step": 94450 | |
| }, | |
| { | |
| "epoch": 27.503318584070797, | |
| "grad_norm": 0.44700610637664795, | |
| "learning_rate": 0.0002701258008153757, | |
| "loss": 3.1676, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 27.517873777363764, | |
| "grad_norm": 0.45396730303764343, | |
| "learning_rate": 0.0002699510774606872, | |
| "loss": 3.1596, | |
| "step": 94550 | |
| }, | |
| { | |
| "epoch": 27.53242897065673, | |
| "grad_norm": 0.41291770339012146, | |
| "learning_rate": 0.0002697763541059988, | |
| "loss": 3.1769, | |
| "step": 94600 | |
| }, | |
| { | |
| "epoch": 27.5469841639497, | |
| "grad_norm": 0.44958990812301636, | |
| "learning_rate": 0.0002696016307513104, | |
| "loss": 3.1754, | |
| "step": 94650 | |
| }, | |
| { | |
| "epoch": 27.561539357242665, | |
| "grad_norm": 0.4432903528213501, | |
| "learning_rate": 0.000269426907396622, | |
| "loss": 3.165, | |
| "step": 94700 | |
| }, | |
| { | |
| "epoch": 27.576094550535633, | |
| "grad_norm": 0.43412747979164124, | |
| "learning_rate": 0.0002692521840419336, | |
| "loss": 3.1682, | |
| "step": 94750 | |
| }, | |
| { | |
| "epoch": 27.5906497438286, | |
| "grad_norm": 0.4156794846057892, | |
| "learning_rate": 0.0002690774606872452, | |
| "loss": 3.1573, | |
| "step": 94800 | |
| }, | |
| { | |
| "epoch": 27.605204937121567, | |
| "grad_norm": 0.4606965184211731, | |
| "learning_rate": 0.0002689027373325568, | |
| "loss": 3.1712, | |
| "step": 94850 | |
| }, | |
| { | |
| "epoch": 27.619760130414534, | |
| "grad_norm": 0.4299561083316803, | |
| "learning_rate": 0.0002687280139778683, | |
| "loss": 3.1645, | |
| "step": 94900 | |
| }, | |
| { | |
| "epoch": 27.634315323707497, | |
| "grad_norm": 0.4365476667881012, | |
| "learning_rate": 0.00026855329062317997, | |
| "loss": 3.1718, | |
| "step": 94950 | |
| }, | |
| { | |
| "epoch": 27.648870517000464, | |
| "grad_norm": 0.4401813745498657, | |
| "learning_rate": 0.0002683785672684915, | |
| "loss": 3.1684, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.648870517000464, | |
| "eval_accuracy": 0.3745749184272355, | |
| "eval_loss": 3.541621446609497, | |
| "eval_runtime": 178.674, | |
| "eval_samples_per_second": 93.192, | |
| "eval_steps_per_second": 5.826, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.66342571029343, | |
| "grad_norm": 0.44295379519462585, | |
| "learning_rate": 0.0002682038439138031, | |
| "loss": 3.1729, | |
| "step": 95050 | |
| }, | |
| { | |
| "epoch": 27.6779809035864, | |
| "grad_norm": 0.4356094300746918, | |
| "learning_rate": 0.0002680291205591147, | |
| "loss": 3.1698, | |
| "step": 95100 | |
| }, | |
| { | |
| "epoch": 27.692536096879365, | |
| "grad_norm": 0.42831119894981384, | |
| "learning_rate": 0.0002678543972044263, | |
| "loss": 3.1754, | |
| "step": 95150 | |
| }, | |
| { | |
| "epoch": 27.707091290172333, | |
| "grad_norm": 0.4276999831199646, | |
| "learning_rate": 0.0002676796738497379, | |
| "loss": 3.1777, | |
| "step": 95200 | |
| }, | |
| { | |
| "epoch": 27.7216464834653, | |
| "grad_norm": 0.42521238327026367, | |
| "learning_rate": 0.0002675049504950495, | |
| "loss": 3.1809, | |
| "step": 95250 | |
| }, | |
| { | |
| "epoch": 27.736201676758267, | |
| "grad_norm": 0.4228277802467346, | |
| "learning_rate": 0.0002673302271403611, | |
| "loss": 3.1755, | |
| "step": 95300 | |
| }, | |
| { | |
| "epoch": 27.750756870051234, | |
| "grad_norm": 0.4252335727214813, | |
| "learning_rate": 0.00026715550378567267, | |
| "loss": 3.1874, | |
| "step": 95350 | |
| }, | |
| { | |
| "epoch": 27.7653120633442, | |
| "grad_norm": 0.5082770586013794, | |
| "learning_rate": 0.00026698078043098426, | |
| "loss": 3.1646, | |
| "step": 95400 | |
| }, | |
| { | |
| "epoch": 27.779867256637168, | |
| "grad_norm": 0.44211241602897644, | |
| "learning_rate": 0.00026680605707629586, | |
| "loss": 3.1714, | |
| "step": 95450 | |
| }, | |
| { | |
| "epoch": 27.794422449930135, | |
| "grad_norm": 0.4420224726200104, | |
| "learning_rate": 0.00026663133372160745, | |
| "loss": 3.1849, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 27.808977643223102, | |
| "grad_norm": 0.46785563230514526, | |
| "learning_rate": 0.00026645661036691905, | |
| "loss": 3.1791, | |
| "step": 95550 | |
| }, | |
| { | |
| "epoch": 27.82353283651607, | |
| "grad_norm": 0.42737874388694763, | |
| "learning_rate": 0.0002662818870122306, | |
| "loss": 3.1751, | |
| "step": 95600 | |
| }, | |
| { | |
| "epoch": 27.838088029809036, | |
| "grad_norm": 0.4250100255012512, | |
| "learning_rate": 0.00026610716365754224, | |
| "loss": 3.1702, | |
| "step": 95650 | |
| }, | |
| { | |
| "epoch": 27.852643223102003, | |
| "grad_norm": 0.44536617398262024, | |
| "learning_rate": 0.0002659324403028538, | |
| "loss": 3.1824, | |
| "step": 95700 | |
| }, | |
| { | |
| "epoch": 27.86719841639497, | |
| "grad_norm": 0.433122843503952, | |
| "learning_rate": 0.00026575771694816537, | |
| "loss": 3.1877, | |
| "step": 95750 | |
| }, | |
| { | |
| "epoch": 27.881753609687937, | |
| "grad_norm": 0.4616224765777588, | |
| "learning_rate": 0.00026558299359347696, | |
| "loss": 3.1831, | |
| "step": 95800 | |
| }, | |
| { | |
| "epoch": 27.896308802980904, | |
| "grad_norm": 0.41990095376968384, | |
| "learning_rate": 0.00026540827023878856, | |
| "loss": 3.1879, | |
| "step": 95850 | |
| }, | |
| { | |
| "epoch": 27.91086399627387, | |
| "grad_norm": 0.4435485303401947, | |
| "learning_rate": 0.00026523354688410015, | |
| "loss": 3.1814, | |
| "step": 95900 | |
| }, | |
| { | |
| "epoch": 27.92541918956684, | |
| "grad_norm": 0.43030399084091187, | |
| "learning_rate": 0.00026505882352941175, | |
| "loss": 3.2003, | |
| "step": 95950 | |
| }, | |
| { | |
| "epoch": 27.939974382859806, | |
| "grad_norm": 0.42241382598876953, | |
| "learning_rate": 0.00026488410017472334, | |
| "loss": 3.1868, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.939974382859806, | |
| "eval_accuracy": 0.375420879971596, | |
| "eval_loss": 3.5337677001953125, | |
| "eval_runtime": 178.7437, | |
| "eval_samples_per_second": 93.156, | |
| "eval_steps_per_second": 5.824, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.954529576152773, | |
| "grad_norm": 0.46110999584198, | |
| "learning_rate": 0.00026470937682003494, | |
| "loss": 3.1818, | |
| "step": 96050 | |
| }, | |
| { | |
| "epoch": 27.96908476944574, | |
| "grad_norm": 0.4265202283859253, | |
| "learning_rate": 0.00026453465346534653, | |
| "loss": 3.1885, | |
| "step": 96100 | |
| }, | |
| { | |
| "epoch": 27.983639962738707, | |
| "grad_norm": 0.45619866251945496, | |
| "learning_rate": 0.00026435993011065807, | |
| "loss": 3.1869, | |
| "step": 96150 | |
| }, | |
| { | |
| "epoch": 27.998195156031674, | |
| "grad_norm": 0.43252936005592346, | |
| "learning_rate": 0.0002641852067559697, | |
| "loss": 3.1825, | |
| "step": 96200 | |
| }, | |
| { | |
| "epoch": 28.01251746623195, | |
| "grad_norm": 0.45363131165504456, | |
| "learning_rate": 0.00026401048340128126, | |
| "loss": 3.0828, | |
| "step": 96250 | |
| }, | |
| { | |
| "epoch": 28.02707265952492, | |
| "grad_norm": 0.45988988876342773, | |
| "learning_rate": 0.00026383576004659285, | |
| "loss": 3.0949, | |
| "step": 96300 | |
| }, | |
| { | |
| "epoch": 28.041627852817886, | |
| "grad_norm": 0.4816953241825104, | |
| "learning_rate": 0.00026366103669190445, | |
| "loss": 3.0957, | |
| "step": 96350 | |
| }, | |
| { | |
| "epoch": 28.056183046110853, | |
| "grad_norm": 0.4343043267726898, | |
| "learning_rate": 0.00026348631333721604, | |
| "loss": 3.1143, | |
| "step": 96400 | |
| }, | |
| { | |
| "epoch": 28.07073823940382, | |
| "grad_norm": 0.4606913626194, | |
| "learning_rate": 0.00026331158998252764, | |
| "loss": 3.1016, | |
| "step": 96450 | |
| }, | |
| { | |
| "epoch": 28.085293432696787, | |
| "grad_norm": 0.44795188307762146, | |
| "learning_rate": 0.00026313686662783923, | |
| "loss": 3.098, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 28.099848625989754, | |
| "grad_norm": 0.4226738214492798, | |
| "learning_rate": 0.0002629621432731508, | |
| "loss": 3.1013, | |
| "step": 96550 | |
| }, | |
| { | |
| "epoch": 28.11440381928272, | |
| "grad_norm": 0.44529786705970764, | |
| "learning_rate": 0.0002627874199184624, | |
| "loss": 3.1166, | |
| "step": 96600 | |
| }, | |
| { | |
| "epoch": 28.128959012575688, | |
| "grad_norm": 0.43973055481910706, | |
| "learning_rate": 0.000262612696563774, | |
| "loss": 3.103, | |
| "step": 96650 | |
| }, | |
| { | |
| "epoch": 28.143514205868655, | |
| "grad_norm": 0.4689570963382721, | |
| "learning_rate": 0.0002624379732090856, | |
| "loss": 3.1211, | |
| "step": 96700 | |
| }, | |
| { | |
| "epoch": 28.158069399161622, | |
| "grad_norm": 0.42313089966773987, | |
| "learning_rate": 0.0002622632498543972, | |
| "loss": 3.112, | |
| "step": 96750 | |
| }, | |
| { | |
| "epoch": 28.17262459245459, | |
| "grad_norm": 0.5010049939155579, | |
| "learning_rate": 0.0002620885264997088, | |
| "loss": 3.1209, | |
| "step": 96800 | |
| }, | |
| { | |
| "epoch": 28.187179785747556, | |
| "grad_norm": 0.4199390709400177, | |
| "learning_rate": 0.00026191380314502034, | |
| "loss": 3.1294, | |
| "step": 96850 | |
| }, | |
| { | |
| "epoch": 28.201734979040523, | |
| "grad_norm": 0.427876353263855, | |
| "learning_rate": 0.000261739079790332, | |
| "loss": 3.1027, | |
| "step": 96900 | |
| }, | |
| { | |
| "epoch": 28.216290172333487, | |
| "grad_norm": 0.4426726996898651, | |
| "learning_rate": 0.0002615643564356435, | |
| "loss": 3.1277, | |
| "step": 96950 | |
| }, | |
| { | |
| "epoch": 28.230845365626454, | |
| "grad_norm": 0.43064725399017334, | |
| "learning_rate": 0.0002613896330809551, | |
| "loss": 3.1243, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.230845365626454, | |
| "eval_accuracy": 0.37407354694234496, | |
| "eval_loss": 3.555612325668335, | |
| "eval_runtime": 178.5317, | |
| "eval_samples_per_second": 93.266, | |
| "eval_steps_per_second": 5.831, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.24540055891942, | |
| "grad_norm": 0.4461931884288788, | |
| "learning_rate": 0.0002612149097262667, | |
| "loss": 3.1323, | |
| "step": 97050 | |
| }, | |
| { | |
| "epoch": 28.259955752212388, | |
| "grad_norm": 0.43096446990966797, | |
| "learning_rate": 0.0002610401863715783, | |
| "loss": 3.1194, | |
| "step": 97100 | |
| }, | |
| { | |
| "epoch": 28.274510945505355, | |
| "grad_norm": 0.47497278451919556, | |
| "learning_rate": 0.0002608654630168899, | |
| "loss": 3.1272, | |
| "step": 97150 | |
| }, | |
| { | |
| "epoch": 28.289066138798322, | |
| "grad_norm": 0.4501521587371826, | |
| "learning_rate": 0.0002606907396622015, | |
| "loss": 3.1414, | |
| "step": 97200 | |
| }, | |
| { | |
| "epoch": 28.30362133209129, | |
| "grad_norm": 0.43721261620521545, | |
| "learning_rate": 0.0002605160163075131, | |
| "loss": 3.125, | |
| "step": 97250 | |
| }, | |
| { | |
| "epoch": 28.318176525384256, | |
| "grad_norm": 0.4478330612182617, | |
| "learning_rate": 0.0002603412929528247, | |
| "loss": 3.1445, | |
| "step": 97300 | |
| }, | |
| { | |
| "epoch": 28.332731718677223, | |
| "grad_norm": 0.5095174312591553, | |
| "learning_rate": 0.0002601665695981363, | |
| "loss": 3.1409, | |
| "step": 97350 | |
| }, | |
| { | |
| "epoch": 28.34728691197019, | |
| "grad_norm": 0.42724117636680603, | |
| "learning_rate": 0.0002599918462434478, | |
| "loss": 3.1428, | |
| "step": 97400 | |
| }, | |
| { | |
| "epoch": 28.361842105263158, | |
| "grad_norm": 0.4426712989807129, | |
| "learning_rate": 0.00025981712288875947, | |
| "loss": 3.1387, | |
| "step": 97450 | |
| }, | |
| { | |
| "epoch": 28.376397298556125, | |
| "grad_norm": 0.4426730275154114, | |
| "learning_rate": 0.000259642399534071, | |
| "loss": 3.1511, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 28.39095249184909, | |
| "grad_norm": 0.42980238795280457, | |
| "learning_rate": 0.0002594676761793826, | |
| "loss": 3.1457, | |
| "step": 97550 | |
| }, | |
| { | |
| "epoch": 28.40550768514206, | |
| "grad_norm": 0.4393332302570343, | |
| "learning_rate": 0.0002592929528246942, | |
| "loss": 3.1492, | |
| "step": 97600 | |
| }, | |
| { | |
| "epoch": 28.420062878435026, | |
| "grad_norm": 0.43661534786224365, | |
| "learning_rate": 0.0002591182294700058, | |
| "loss": 3.1473, | |
| "step": 97650 | |
| }, | |
| { | |
| "epoch": 28.434618071727993, | |
| "grad_norm": 0.43180370330810547, | |
| "learning_rate": 0.0002589435061153174, | |
| "loss": 3.1527, | |
| "step": 97700 | |
| }, | |
| { | |
| "epoch": 28.44917326502096, | |
| "grad_norm": 0.45117655396461487, | |
| "learning_rate": 0.000258768782760629, | |
| "loss": 3.1472, | |
| "step": 97750 | |
| }, | |
| { | |
| "epoch": 28.463728458313927, | |
| "grad_norm": 0.4737044870853424, | |
| "learning_rate": 0.0002585940594059406, | |
| "loss": 3.1518, | |
| "step": 97800 | |
| }, | |
| { | |
| "epoch": 28.478283651606894, | |
| "grad_norm": 0.4503737688064575, | |
| "learning_rate": 0.00025841933605125217, | |
| "loss": 3.1522, | |
| "step": 97850 | |
| }, | |
| { | |
| "epoch": 28.49283884489986, | |
| "grad_norm": 0.4357546865940094, | |
| "learning_rate": 0.00025824461269656376, | |
| "loss": 3.163, | |
| "step": 97900 | |
| }, | |
| { | |
| "epoch": 28.507394038192828, | |
| "grad_norm": 0.4459472596645355, | |
| "learning_rate": 0.00025806988934187536, | |
| "loss": 3.148, | |
| "step": 97950 | |
| }, | |
| { | |
| "epoch": 28.521949231485795, | |
| "grad_norm": 0.4553315043449402, | |
| "learning_rate": 0.00025789516598718695, | |
| "loss": 3.149, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.521949231485795, | |
| "eval_accuracy": 0.3750310418995421, | |
| "eval_loss": 3.5434606075286865, | |
| "eval_runtime": 178.8143, | |
| "eval_samples_per_second": 93.119, | |
| "eval_steps_per_second": 5.822, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.536504424778762, | |
| "grad_norm": 0.45329025387763977, | |
| "learning_rate": 0.00025772044263249854, | |
| "loss": 3.1607, | |
| "step": 98050 | |
| }, | |
| { | |
| "epoch": 28.55105961807173, | |
| "grad_norm": 0.4378615617752075, | |
| "learning_rate": 0.0002575457192778101, | |
| "loss": 3.1634, | |
| "step": 98100 | |
| }, | |
| { | |
| "epoch": 28.565614811364696, | |
| "grad_norm": 0.4968438446521759, | |
| "learning_rate": 0.00025737099592312173, | |
| "loss": 3.1435, | |
| "step": 98150 | |
| }, | |
| { | |
| "epoch": 28.580170004657663, | |
| "grad_norm": 0.4444005787372589, | |
| "learning_rate": 0.0002571962725684333, | |
| "loss": 3.1574, | |
| "step": 98200 | |
| }, | |
| { | |
| "epoch": 28.59472519795063, | |
| "grad_norm": 0.4303135275840759, | |
| "learning_rate": 0.00025702154921374487, | |
| "loss": 3.1583, | |
| "step": 98250 | |
| }, | |
| { | |
| "epoch": 28.609280391243594, | |
| "grad_norm": 0.4470735192298889, | |
| "learning_rate": 0.00025684682585905646, | |
| "loss": 3.1603, | |
| "step": 98300 | |
| }, | |
| { | |
| "epoch": 28.62383558453656, | |
| "grad_norm": 0.44142308831214905, | |
| "learning_rate": 0.00025667210250436806, | |
| "loss": 3.1726, | |
| "step": 98350 | |
| }, | |
| { | |
| "epoch": 28.638390777829528, | |
| "grad_norm": 0.46621376276016235, | |
| "learning_rate": 0.00025649737914967965, | |
| "loss": 3.1674, | |
| "step": 98400 | |
| }, | |
| { | |
| "epoch": 28.652945971122495, | |
| "grad_norm": 0.4749205410480499, | |
| "learning_rate": 0.00025632265579499124, | |
| "loss": 3.1734, | |
| "step": 98450 | |
| }, | |
| { | |
| "epoch": 28.667501164415462, | |
| "grad_norm": 0.4264492988586426, | |
| "learning_rate": 0.00025614793244030284, | |
| "loss": 3.1614, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 28.68205635770843, | |
| "grad_norm": 0.8595496416091919, | |
| "learning_rate": 0.0002559732090856144, | |
| "loss": 3.1595, | |
| "step": 98550 | |
| }, | |
| { | |
| "epoch": 28.696611551001396, | |
| "grad_norm": 0.45727524161338806, | |
| "learning_rate": 0.00025579848573092603, | |
| "loss": 3.157, | |
| "step": 98600 | |
| }, | |
| { | |
| "epoch": 28.711166744294363, | |
| "grad_norm": 0.4491088390350342, | |
| "learning_rate": 0.0002556237623762376, | |
| "loss": 3.1799, | |
| "step": 98650 | |
| }, | |
| { | |
| "epoch": 28.72572193758733, | |
| "grad_norm": 0.4161786735057831, | |
| "learning_rate": 0.0002554490390215492, | |
| "loss": 3.1753, | |
| "step": 98700 | |
| }, | |
| { | |
| "epoch": 28.740277130880298, | |
| "grad_norm": 0.4785554111003876, | |
| "learning_rate": 0.0002552743156668608, | |
| "loss": 3.1759, | |
| "step": 98750 | |
| }, | |
| { | |
| "epoch": 28.754832324173265, | |
| "grad_norm": 0.4457080066204071, | |
| "learning_rate": 0.00025509959231217235, | |
| "loss": 3.1696, | |
| "step": 98800 | |
| }, | |
| { | |
| "epoch": 28.76938751746623, | |
| "grad_norm": 0.4306055009365082, | |
| "learning_rate": 0.000254924868957484, | |
| "loss": 3.1665, | |
| "step": 98850 | |
| }, | |
| { | |
| "epoch": 28.7839427107592, | |
| "grad_norm": 0.4257147014141083, | |
| "learning_rate": 0.00025475014560279554, | |
| "loss": 3.1832, | |
| "step": 98900 | |
| }, | |
| { | |
| "epoch": 28.798497904052166, | |
| "grad_norm": 0.4394259452819824, | |
| "learning_rate": 0.00025457542224810713, | |
| "loss": 3.1635, | |
| "step": 98950 | |
| }, | |
| { | |
| "epoch": 28.813053097345133, | |
| "grad_norm": 0.43619710206985474, | |
| "learning_rate": 0.00025440069889341873, | |
| "loss": 3.1667, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.813053097345133, | |
| "eval_accuracy": 0.3752215536616161, | |
| "eval_loss": 3.5366761684417725, | |
| "eval_runtime": 178.9311, | |
| "eval_samples_per_second": 93.058, | |
| "eval_steps_per_second": 5.818, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.8276082906381, | |
| "grad_norm": 0.46083733439445496, | |
| "learning_rate": 0.0002542259755387303, | |
| "loss": 3.1623, | |
| "step": 99050 | |
| }, | |
| { | |
| "epoch": 28.842163483931067, | |
| "grad_norm": 0.48397934436798096, | |
| "learning_rate": 0.0002540512521840419, | |
| "loss": 3.1784, | |
| "step": 99100 | |
| }, | |
| { | |
| "epoch": 28.856718677224034, | |
| "grad_norm": 0.456592321395874, | |
| "learning_rate": 0.0002538765288293535, | |
| "loss": 3.1693, | |
| "step": 99150 | |
| }, | |
| { | |
| "epoch": 28.871273870517, | |
| "grad_norm": 0.4616560637950897, | |
| "learning_rate": 0.0002537018054746651, | |
| "loss": 3.1686, | |
| "step": 99200 | |
| }, | |
| { | |
| "epoch": 28.885829063809968, | |
| "grad_norm": 0.44148775935173035, | |
| "learning_rate": 0.00025352708211997664, | |
| "loss": 3.1726, | |
| "step": 99250 | |
| }, | |
| { | |
| "epoch": 28.900384257102935, | |
| "grad_norm": 0.4602034389972687, | |
| "learning_rate": 0.0002533523587652883, | |
| "loss": 3.1717, | |
| "step": 99300 | |
| }, | |
| { | |
| "epoch": 28.914939450395902, | |
| "grad_norm": 0.4577277600765228, | |
| "learning_rate": 0.00025317763541059983, | |
| "loss": 3.1874, | |
| "step": 99350 | |
| }, | |
| { | |
| "epoch": 28.92949464368887, | |
| "grad_norm": 0.46108654141426086, | |
| "learning_rate": 0.0002530029120559115, | |
| "loss": 3.1752, | |
| "step": 99400 | |
| }, | |
| { | |
| "epoch": 28.944049836981836, | |
| "grad_norm": 0.4398329555988312, | |
| "learning_rate": 0.000252828188701223, | |
| "loss": 3.1825, | |
| "step": 99450 | |
| }, | |
| { | |
| "epoch": 28.958605030274803, | |
| "grad_norm": 0.4647148847579956, | |
| "learning_rate": 0.0002526534653465346, | |
| "loss": 3.1825, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 28.97316022356777, | |
| "grad_norm": 0.48289281129837036, | |
| "learning_rate": 0.0002524787419918462, | |
| "loss": 3.1766, | |
| "step": 99550 | |
| }, | |
| { | |
| "epoch": 28.987715416860738, | |
| "grad_norm": 0.4246104955673218, | |
| "learning_rate": 0.0002523040186371578, | |
| "loss": 3.179, | |
| "step": 99600 | |
| }, | |
| { | |
| "epoch": 29.002037727061015, | |
| "grad_norm": 0.4434986412525177, | |
| "learning_rate": 0.0002521292952824694, | |
| "loss": 3.1789, | |
| "step": 99650 | |
| }, | |
| { | |
| "epoch": 29.016592920353983, | |
| "grad_norm": 0.481111615896225, | |
| "learning_rate": 0.000251954571927781, | |
| "loss": 3.0767, | |
| "step": 99700 | |
| }, | |
| { | |
| "epoch": 29.03114811364695, | |
| "grad_norm": 0.4342561960220337, | |
| "learning_rate": 0.0002517798485730926, | |
| "loss": 3.0848, | |
| "step": 99750 | |
| }, | |
| { | |
| "epoch": 29.045703306939917, | |
| "grad_norm": 0.4534182548522949, | |
| "learning_rate": 0.0002516051252184042, | |
| "loss": 3.08, | |
| "step": 99800 | |
| }, | |
| { | |
| "epoch": 29.060258500232884, | |
| "grad_norm": 0.4300639033317566, | |
| "learning_rate": 0.0002514304018637158, | |
| "loss": 3.0883, | |
| "step": 99850 | |
| }, | |
| { | |
| "epoch": 29.07481369352585, | |
| "grad_norm": 0.43499693274497986, | |
| "learning_rate": 0.00025125567850902737, | |
| "loss": 3.0863, | |
| "step": 99900 | |
| }, | |
| { | |
| "epoch": 29.089368886818818, | |
| "grad_norm": 0.4732967019081116, | |
| "learning_rate": 0.0002510809551543389, | |
| "loss": 3.0884, | |
| "step": 99950 | |
| }, | |
| { | |
| "epoch": 29.103924080111785, | |
| "grad_norm": 0.441400945186615, | |
| "learning_rate": 0.00025090623179965056, | |
| "loss": 3.1054, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 29.103924080111785, | |
| "eval_accuracy": 0.37428838685663934, | |
| "eval_loss": 3.5510168075561523, | |
| "eval_runtime": 186.9219, | |
| "eval_samples_per_second": 89.08, | |
| "eval_steps_per_second": 5.569, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 29.118479273404752, | |
| "grad_norm": 0.44024235010147095, | |
| "learning_rate": 0.0002507315084449621, | |
| "loss": 3.111, | |
| "step": 100050 | |
| }, | |
| { | |
| "epoch": 29.13303446669772, | |
| "grad_norm": 0.4516540765762329, | |
| "learning_rate": 0.00025055678509027375, | |
| "loss": 3.092, | |
| "step": 100100 | |
| }, | |
| { | |
| "epoch": 29.147589659990686, | |
| "grad_norm": 0.43763741850852966, | |
| "learning_rate": 0.0002503820617355853, | |
| "loss": 3.1217, | |
| "step": 100150 | |
| }, | |
| { | |
| "epoch": 29.162144853283653, | |
| "grad_norm": 0.4587101638317108, | |
| "learning_rate": 0.0002502073383808969, | |
| "loss": 3.1219, | |
| "step": 100200 | |
| }, | |
| { | |
| "epoch": 29.17670004657662, | |
| "grad_norm": 0.4364701807498932, | |
| "learning_rate": 0.0002500326150262085, | |
| "loss": 3.1208, | |
| "step": 100250 | |
| }, | |
| { | |
| "epoch": 29.191255239869584, | |
| "grad_norm": 0.4533688426017761, | |
| "learning_rate": 0.00024985789167152007, | |
| "loss": 3.117, | |
| "step": 100300 | |
| }, | |
| { | |
| "epoch": 29.20581043316255, | |
| "grad_norm": 0.454554945230484, | |
| "learning_rate": 0.00024968316831683167, | |
| "loss": 3.1305, | |
| "step": 100350 | |
| }, | |
| { | |
| "epoch": 29.220365626455518, | |
| "grad_norm": 0.43322986364364624, | |
| "learning_rate": 0.00024950844496214326, | |
| "loss": 3.1151, | |
| "step": 100400 | |
| }, | |
| { | |
| "epoch": 29.234920819748485, | |
| "grad_norm": 0.4622367322444916, | |
| "learning_rate": 0.00024933372160745485, | |
| "loss": 3.1159, | |
| "step": 100450 | |
| }, | |
| { | |
| "epoch": 29.249476013041452, | |
| "grad_norm": 0.4399924874305725, | |
| "learning_rate": 0.0002491589982527664, | |
| "loss": 3.125, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 29.26403120633442, | |
| "grad_norm": 0.4242802560329437, | |
| "learning_rate": 0.00024898427489807804, | |
| "loss": 3.111, | |
| "step": 100550 | |
| }, | |
| { | |
| "epoch": 29.278586399627386, | |
| "grad_norm": 0.46443307399749756, | |
| "learning_rate": 0.0002488095515433896, | |
| "loss": 3.1034, | |
| "step": 100600 | |
| }, | |
| { | |
| "epoch": 29.293141592920353, | |
| "grad_norm": 0.4604977071285248, | |
| "learning_rate": 0.0002486348281887012, | |
| "loss": 3.1327, | |
| "step": 100650 | |
| }, | |
| { | |
| "epoch": 29.30769678621332, | |
| "grad_norm": 0.45278415083885193, | |
| "learning_rate": 0.00024846010483401277, | |
| "loss": 3.1281, | |
| "step": 100700 | |
| }, | |
| { | |
| "epoch": 29.322251979506287, | |
| "grad_norm": 0.4434463381767273, | |
| "learning_rate": 0.00024828538147932437, | |
| "loss": 3.1237, | |
| "step": 100750 | |
| }, | |
| { | |
| "epoch": 29.336807172799254, | |
| "grad_norm": 0.435261607170105, | |
| "learning_rate": 0.00024811065812463596, | |
| "loss": 3.1243, | |
| "step": 100800 | |
| }, | |
| { | |
| "epoch": 29.35136236609222, | |
| "grad_norm": 0.4782884418964386, | |
| "learning_rate": 0.00024793593476994755, | |
| "loss": 3.1301, | |
| "step": 100850 | |
| }, | |
| { | |
| "epoch": 29.36591755938519, | |
| "grad_norm": 0.4731723368167877, | |
| "learning_rate": 0.00024776121141525915, | |
| "loss": 3.1214, | |
| "step": 100900 | |
| }, | |
| { | |
| "epoch": 29.380472752678156, | |
| "grad_norm": 0.4642447233200073, | |
| "learning_rate": 0.00024758648806057074, | |
| "loss": 3.1421, | |
| "step": 100950 | |
| }, | |
| { | |
| "epoch": 29.395027945971123, | |
| "grad_norm": 0.4479381740093231, | |
| "learning_rate": 0.00024741176470588234, | |
| "loss": 3.1242, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 29.395027945971123, | |
| "eval_accuracy": 0.3747103098830709, | |
| "eval_loss": 3.5458545684814453, | |
| "eval_runtime": 179.2165, | |
| "eval_samples_per_second": 92.91, | |
| "eval_steps_per_second": 5.809, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 29.40958313926409, | |
| "grad_norm": 0.4384738802909851, | |
| "learning_rate": 0.00024723704135119393, | |
| "loss": 3.1327, | |
| "step": 101050 | |
| }, | |
| { | |
| "epoch": 29.424138332557057, | |
| "grad_norm": 0.436471551656723, | |
| "learning_rate": 0.0002470623179965055, | |
| "loss": 3.1351, | |
| "step": 101100 | |
| }, | |
| { | |
| "epoch": 29.438693525850024, | |
| "grad_norm": 0.4630633294582367, | |
| "learning_rate": 0.0002468875946418171, | |
| "loss": 3.1315, | |
| "step": 101150 | |
| }, | |
| { | |
| "epoch": 29.45324871914299, | |
| "grad_norm": 0.43443524837493896, | |
| "learning_rate": 0.00024671287128712866, | |
| "loss": 3.153, | |
| "step": 101200 | |
| }, | |
| { | |
| "epoch": 29.467803912435958, | |
| "grad_norm": 0.5008822083473206, | |
| "learning_rate": 0.0002465381479324403, | |
| "loss": 3.142, | |
| "step": 101250 | |
| }, | |
| { | |
| "epoch": 29.482359105728925, | |
| "grad_norm": 0.4582183063030243, | |
| "learning_rate": 0.00024636342457775185, | |
| "loss": 3.155, | |
| "step": 101300 | |
| }, | |
| { | |
| "epoch": 29.496914299021892, | |
| "grad_norm": 0.42436885833740234, | |
| "learning_rate": 0.00024618870122306344, | |
| "loss": 3.1489, | |
| "step": 101350 | |
| }, | |
| { | |
| "epoch": 29.51146949231486, | |
| "grad_norm": 0.475890189409256, | |
| "learning_rate": 0.00024601397786837504, | |
| "loss": 3.1403, | |
| "step": 101400 | |
| }, | |
| { | |
| "epoch": 29.526024685607826, | |
| "grad_norm": 0.43587788939476013, | |
| "learning_rate": 0.00024583925451368663, | |
| "loss": 3.134, | |
| "step": 101450 | |
| }, | |
| { | |
| "epoch": 29.540579878900793, | |
| "grad_norm": 0.45687776803970337, | |
| "learning_rate": 0.0002456645311589982, | |
| "loss": 3.1401, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 29.55513507219376, | |
| "grad_norm": 0.47364842891693115, | |
| "learning_rate": 0.0002454898078043098, | |
| "loss": 3.1547, | |
| "step": 101550 | |
| }, | |
| { | |
| "epoch": 29.569690265486727, | |
| "grad_norm": 0.46195733547210693, | |
| "learning_rate": 0.0002453150844496214, | |
| "loss": 3.1403, | |
| "step": 101600 | |
| }, | |
| { | |
| "epoch": 29.58424545877969, | |
| "grad_norm": 0.4433932304382324, | |
| "learning_rate": 0.000245140361094933, | |
| "loss": 3.1476, | |
| "step": 101650 | |
| }, | |
| { | |
| "epoch": 29.598800652072658, | |
| "grad_norm": 0.4796285927295685, | |
| "learning_rate": 0.0002449656377402446, | |
| "loss": 3.1543, | |
| "step": 101700 | |
| }, | |
| { | |
| "epoch": 29.613355845365625, | |
| "grad_norm": 0.4393634498119354, | |
| "learning_rate": 0.00024479091438555614, | |
| "loss": 3.1608, | |
| "step": 101750 | |
| }, | |
| { | |
| "epoch": 29.627911038658592, | |
| "grad_norm": 0.4540051519870758, | |
| "learning_rate": 0.0002446161910308678, | |
| "loss": 3.1679, | |
| "step": 101800 | |
| }, | |
| { | |
| "epoch": 29.64246623195156, | |
| "grad_norm": 0.45071864128112793, | |
| "learning_rate": 0.0002444414676761794, | |
| "loss": 3.1577, | |
| "step": 101850 | |
| }, | |
| { | |
| "epoch": 29.657021425244526, | |
| "grad_norm": 0.44411563873291016, | |
| "learning_rate": 0.0002442667443214909, | |
| "loss": 3.1719, | |
| "step": 101900 | |
| }, | |
| { | |
| "epoch": 29.671576618537493, | |
| "grad_norm": 0.452627956867218, | |
| "learning_rate": 0.00024409202096680255, | |
| "loss": 3.1593, | |
| "step": 101950 | |
| }, | |
| { | |
| "epoch": 29.68613181183046, | |
| "grad_norm": 0.4586232602596283, | |
| "learning_rate": 0.00024391729761211411, | |
| "loss": 3.168, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 29.68613181183046, | |
| "eval_accuracy": 0.37510555421117375, | |
| "eval_loss": 3.5422708988189697, | |
| "eval_runtime": 179.3978, | |
| "eval_samples_per_second": 92.816, | |
| "eval_steps_per_second": 5.803, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 29.700687005123427, | |
| "grad_norm": 0.44371819496154785, | |
| "learning_rate": 0.00024374257425742574, | |
| "loss": 3.159, | |
| "step": 102050 | |
| }, | |
| { | |
| "epoch": 29.715242198416394, | |
| "grad_norm": 0.4582088589668274, | |
| "learning_rate": 0.0002435678509027373, | |
| "loss": 3.1497, | |
| "step": 102100 | |
| }, | |
| { | |
| "epoch": 29.72979739170936, | |
| "grad_norm": 0.467153936624527, | |
| "learning_rate": 0.0002433931275480489, | |
| "loss": 3.1633, | |
| "step": 102150 | |
| }, | |
| { | |
| "epoch": 29.74435258500233, | |
| "grad_norm": 0.4523339867591858, | |
| "learning_rate": 0.00024321840419336052, | |
| "loss": 3.1521, | |
| "step": 102200 | |
| }, | |
| { | |
| "epoch": 29.758907778295296, | |
| "grad_norm": 0.4426189959049225, | |
| "learning_rate": 0.00024304368083867209, | |
| "loss": 3.1679, | |
| "step": 102250 | |
| }, | |
| { | |
| "epoch": 29.773462971588263, | |
| "grad_norm": 0.43963423371315, | |
| "learning_rate": 0.00024286895748398365, | |
| "loss": 3.1501, | |
| "step": 102300 | |
| }, | |
| { | |
| "epoch": 29.78801816488123, | |
| "grad_norm": 0.4491553008556366, | |
| "learning_rate": 0.00024269423412929527, | |
| "loss": 3.1566, | |
| "step": 102350 | |
| }, | |
| { | |
| "epoch": 29.802573358174197, | |
| "grad_norm": 0.46082642674446106, | |
| "learning_rate": 0.00024251951077460684, | |
| "loss": 3.1595, | |
| "step": 102400 | |
| }, | |
| { | |
| "epoch": 29.817128551467164, | |
| "grad_norm": 0.4386240541934967, | |
| "learning_rate": 0.00024234478741991844, | |
| "loss": 3.172, | |
| "step": 102450 | |
| }, | |
| { | |
| "epoch": 29.83168374476013, | |
| "grad_norm": 0.4534355700016022, | |
| "learning_rate": 0.00024217006406523003, | |
| "loss": 3.1628, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 29.846238938053098, | |
| "grad_norm": 0.44121938943862915, | |
| "learning_rate": 0.00024199534071054162, | |
| "loss": 3.1749, | |
| "step": 102550 | |
| }, | |
| { | |
| "epoch": 29.860794131346065, | |
| "grad_norm": 0.48554858565330505, | |
| "learning_rate": 0.0002418206173558532, | |
| "loss": 3.1665, | |
| "step": 102600 | |
| }, | |
| { | |
| "epoch": 29.875349324639032, | |
| "grad_norm": 0.45584553480148315, | |
| "learning_rate": 0.0002416458940011648, | |
| "loss": 3.1708, | |
| "step": 102650 | |
| }, | |
| { | |
| "epoch": 29.889904517932, | |
| "grad_norm": 0.44270968437194824, | |
| "learning_rate": 0.00024147117064647638, | |
| "loss": 3.1641, | |
| "step": 102700 | |
| }, | |
| { | |
| "epoch": 29.904459711224966, | |
| "grad_norm": 0.47318506240844727, | |
| "learning_rate": 0.000241296447291788, | |
| "loss": 3.1675, | |
| "step": 102750 | |
| }, | |
| { | |
| "epoch": 29.919014904517933, | |
| "grad_norm": 0.4316175580024719, | |
| "learning_rate": 0.00024112172393709957, | |
| "loss": 3.1721, | |
| "step": 102800 | |
| }, | |
| { | |
| "epoch": 29.9335700978109, | |
| "grad_norm": 0.44327935576438904, | |
| "learning_rate": 0.00024094700058241116, | |
| "loss": 3.1656, | |
| "step": 102850 | |
| }, | |
| { | |
| "epoch": 29.948125291103867, | |
| "grad_norm": 0.44737857580184937, | |
| "learning_rate": 0.00024077227722772276, | |
| "loss": 3.162, | |
| "step": 102900 | |
| }, | |
| { | |
| "epoch": 29.962680484396834, | |
| "grad_norm": 0.4546486735343933, | |
| "learning_rate": 0.00024059755387303435, | |
| "loss": 3.1818, | |
| "step": 102950 | |
| }, | |
| { | |
| "epoch": 29.977235677689798, | |
| "grad_norm": 0.43689993023872375, | |
| "learning_rate": 0.00024042283051834592, | |
| "loss": 3.1654, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 29.977235677689798, | |
| "eval_accuracy": 0.37546554034765284, | |
| "eval_loss": 3.537097454071045, | |
| "eval_runtime": 179.3374, | |
| "eval_samples_per_second": 92.847, | |
| "eval_steps_per_second": 5.805, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 29.991790870982765, | |
| "grad_norm": 0.45586585998535156, | |
| "learning_rate": 0.00024024810716365754, | |
| "loss": 3.1683, | |
| "step": 103050 | |
| }, | |
| { | |
| "epoch": 30.006113181183046, | |
| "grad_norm": 0.4485780596733093, | |
| "learning_rate": 0.0002400733838089691, | |
| "loss": 3.128, | |
| "step": 103100 | |
| }, | |
| { | |
| "epoch": 30.020668374476013, | |
| "grad_norm": 0.4525091350078583, | |
| "learning_rate": 0.0002398986604542807, | |
| "loss": 3.0769, | |
| "step": 103150 | |
| }, | |
| { | |
| "epoch": 30.03522356776898, | |
| "grad_norm": 0.45657244324684143, | |
| "learning_rate": 0.0002397239370995923, | |
| "loss": 3.0872, | |
| "step": 103200 | |
| }, | |
| { | |
| "epoch": 30.049778761061948, | |
| "grad_norm": 0.46582069993019104, | |
| "learning_rate": 0.0002395492137449039, | |
| "loss": 3.0911, | |
| "step": 103250 | |
| }, | |
| { | |
| "epoch": 30.064333954354915, | |
| "grad_norm": 0.47504469752311707, | |
| "learning_rate": 0.00023937449039021546, | |
| "loss": 3.0842, | |
| "step": 103300 | |
| }, | |
| { | |
| "epoch": 30.07888914764788, | |
| "grad_norm": 0.4897063672542572, | |
| "learning_rate": 0.00023919976703552708, | |
| "loss": 3.0813, | |
| "step": 103350 | |
| }, | |
| { | |
| "epoch": 30.09344434094085, | |
| "grad_norm": 0.4772857427597046, | |
| "learning_rate": 0.00023902504368083865, | |
| "loss": 3.0985, | |
| "step": 103400 | |
| }, | |
| { | |
| "epoch": 30.107999534233816, | |
| "grad_norm": 0.43550288677215576, | |
| "learning_rate": 0.00023885032032615027, | |
| "loss": 3.1014, | |
| "step": 103450 | |
| }, | |
| { | |
| "epoch": 30.122554727526783, | |
| "grad_norm": 0.4723055958747864, | |
| "learning_rate": 0.00023867559697146183, | |
| "loss": 3.1032, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 30.13710992081975, | |
| "grad_norm": 0.4526902735233307, | |
| "learning_rate": 0.0002385008736167734, | |
| "loss": 3.0951, | |
| "step": 103550 | |
| }, | |
| { | |
| "epoch": 30.151665114112717, | |
| "grad_norm": 0.4515558183193207, | |
| "learning_rate": 0.00023832615026208502, | |
| "loss": 3.1004, | |
| "step": 103600 | |
| }, | |
| { | |
| "epoch": 30.166220307405684, | |
| "grad_norm": 0.4724433720111847, | |
| "learning_rate": 0.0002381514269073966, | |
| "loss": 3.0963, | |
| "step": 103650 | |
| }, | |
| { | |
| "epoch": 30.180775500698648, | |
| "grad_norm": 0.45959505438804626, | |
| "learning_rate": 0.00023797670355270818, | |
| "loss": 3.1053, | |
| "step": 103700 | |
| }, | |
| { | |
| "epoch": 30.195330693991615, | |
| "grad_norm": 0.47177109122276306, | |
| "learning_rate": 0.0002378019801980198, | |
| "loss": 3.1237, | |
| "step": 103750 | |
| }, | |
| { | |
| "epoch": 30.20988588728458, | |
| "grad_norm": 0.4628247618675232, | |
| "learning_rate": 0.00023762725684333137, | |
| "loss": 3.1103, | |
| "step": 103800 | |
| }, | |
| { | |
| "epoch": 30.22444108057755, | |
| "grad_norm": 0.4638172388076782, | |
| "learning_rate": 0.00023745253348864294, | |
| "loss": 3.1001, | |
| "step": 103850 | |
| }, | |
| { | |
| "epoch": 30.238996273870516, | |
| "grad_norm": 0.4565567672252655, | |
| "learning_rate": 0.00023727781013395456, | |
| "loss": 3.114, | |
| "step": 103900 | |
| }, | |
| { | |
| "epoch": 30.253551467163483, | |
| "grad_norm": 0.44833269715309143, | |
| "learning_rate": 0.00023710308677926613, | |
| "loss": 3.1142, | |
| "step": 103950 | |
| }, | |
| { | |
| "epoch": 30.26810666045645, | |
| "grad_norm": 0.4612325429916382, | |
| "learning_rate": 0.00023692836342457772, | |
| "loss": 3.1044, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 30.26810666045645, | |
| "eval_accuracy": 0.37487837392981105, | |
| "eval_loss": 3.5515880584716797, | |
| "eval_runtime": 179.1596, | |
| "eval_samples_per_second": 92.939, | |
| "eval_steps_per_second": 5.81, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 30.282661853749417, | |
| "grad_norm": 0.46742621064186096, | |
| "learning_rate": 0.00023675364006988932, | |
| "loss": 3.1101, | |
| "step": 104050 | |
| }, | |
| { | |
| "epoch": 30.297217047042384, | |
| "grad_norm": 0.4602166712284088, | |
| "learning_rate": 0.0002365789167152009, | |
| "loss": 3.1211, | |
| "step": 104100 | |
| }, | |
| { | |
| "epoch": 30.31177224033535, | |
| "grad_norm": 0.45198196172714233, | |
| "learning_rate": 0.0002364041933605125, | |
| "loss": 3.1238, | |
| "step": 104150 | |
| }, | |
| { | |
| "epoch": 30.326327433628318, | |
| "grad_norm": 0.44531548023223877, | |
| "learning_rate": 0.0002362294700058241, | |
| "loss": 3.1364, | |
| "step": 104200 | |
| }, | |
| { | |
| "epoch": 30.340882626921285, | |
| "grad_norm": 0.44484877586364746, | |
| "learning_rate": 0.00023605474665113567, | |
| "loss": 3.1233, | |
| "step": 104250 | |
| }, | |
| { | |
| "epoch": 30.355437820214252, | |
| "grad_norm": 0.4585815668106079, | |
| "learning_rate": 0.0002358800232964473, | |
| "loss": 3.117, | |
| "step": 104300 | |
| }, | |
| { | |
| "epoch": 30.36999301350722, | |
| "grad_norm": 0.4720662832260132, | |
| "learning_rate": 0.00023570529994175886, | |
| "loss": 3.1192, | |
| "step": 104350 | |
| }, | |
| { | |
| "epoch": 30.384548206800186, | |
| "grad_norm": 0.4488585293292999, | |
| "learning_rate": 0.00023553057658707045, | |
| "loss": 3.1271, | |
| "step": 104400 | |
| }, | |
| { | |
| "epoch": 30.399103400093153, | |
| "grad_norm": 0.46229711174964905, | |
| "learning_rate": 0.00023535585323238204, | |
| "loss": 3.1308, | |
| "step": 104450 | |
| }, | |
| { | |
| "epoch": 30.41365859338612, | |
| "grad_norm": 0.4812774360179901, | |
| "learning_rate": 0.00023518112987769364, | |
| "loss": 3.1154, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 30.428213786679088, | |
| "grad_norm": 0.45849233865737915, | |
| "learning_rate": 0.0002350064065230052, | |
| "loss": 3.1383, | |
| "step": 104550 | |
| }, | |
| { | |
| "epoch": 30.442768979972055, | |
| "grad_norm": 0.46144720911979675, | |
| "learning_rate": 0.00023483168316831683, | |
| "loss": 3.1309, | |
| "step": 104600 | |
| }, | |
| { | |
| "epoch": 30.45732417326502, | |
| "grad_norm": 0.4718158543109894, | |
| "learning_rate": 0.0002346569598136284, | |
| "loss": 3.1369, | |
| "step": 104650 | |
| }, | |
| { | |
| "epoch": 30.47187936655799, | |
| "grad_norm": 0.4972895383834839, | |
| "learning_rate": 0.00023448223645894, | |
| "loss": 3.1443, | |
| "step": 104700 | |
| }, | |
| { | |
| "epoch": 30.486434559850956, | |
| "grad_norm": 0.44092628359794617, | |
| "learning_rate": 0.00023430751310425158, | |
| "loss": 3.1188, | |
| "step": 104750 | |
| }, | |
| { | |
| "epoch": 30.500989753143923, | |
| "grad_norm": 0.44488775730133057, | |
| "learning_rate": 0.00023413278974956318, | |
| "loss": 3.1493, | |
| "step": 104800 | |
| }, | |
| { | |
| "epoch": 30.51554494643689, | |
| "grad_norm": 0.44548696279525757, | |
| "learning_rate": 0.00023395806639487477, | |
| "loss": 3.1299, | |
| "step": 104850 | |
| }, | |
| { | |
| "epoch": 30.530100139729857, | |
| "grad_norm": 0.4458260238170624, | |
| "learning_rate": 0.00023378334304018637, | |
| "loss": 3.1381, | |
| "step": 104900 | |
| }, | |
| { | |
| "epoch": 30.544655333022824, | |
| "grad_norm": 0.4345245957374573, | |
| "learning_rate": 0.00023360861968549793, | |
| "loss": 3.1361, | |
| "step": 104950 | |
| }, | |
| { | |
| "epoch": 30.55921052631579, | |
| "grad_norm": 0.4373602271080017, | |
| "learning_rate": 0.00023343389633080955, | |
| "loss": 3.135, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 30.55921052631579, | |
| "eval_accuracy": 0.3747435701105027, | |
| "eval_loss": 3.5478925704956055, | |
| "eval_runtime": 179.1076, | |
| "eval_samples_per_second": 92.966, | |
| "eval_steps_per_second": 5.812, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 30.573765719608758, | |
| "grad_norm": 0.47670161724090576, | |
| "learning_rate": 0.00023325917297612112, | |
| "loss": 3.1391, | |
| "step": 105050 | |
| }, | |
| { | |
| "epoch": 30.58832091290172, | |
| "grad_norm": 0.43521973490715027, | |
| "learning_rate": 0.0002330844496214327, | |
| "loss": 3.1376, | |
| "step": 105100 | |
| }, | |
| { | |
| "epoch": 30.60287610619469, | |
| "grad_norm": 0.4696406424045563, | |
| "learning_rate": 0.0002329097262667443, | |
| "loss": 3.1465, | |
| "step": 105150 | |
| }, | |
| { | |
| "epoch": 30.617431299487656, | |
| "grad_norm": 0.43479442596435547, | |
| "learning_rate": 0.00023273500291205588, | |
| "loss": 3.1442, | |
| "step": 105200 | |
| }, | |
| { | |
| "epoch": 30.631986492780623, | |
| "grad_norm": 0.4464789927005768, | |
| "learning_rate": 0.00023256027955736747, | |
| "loss": 3.1509, | |
| "step": 105250 | |
| }, | |
| { | |
| "epoch": 30.64654168607359, | |
| "grad_norm": 0.49761533737182617, | |
| "learning_rate": 0.00023238555620267907, | |
| "loss": 3.1401, | |
| "step": 105300 | |
| }, | |
| { | |
| "epoch": 30.661096879366557, | |
| "grad_norm": 0.4205106496810913, | |
| "learning_rate": 0.00023221083284799066, | |
| "loss": 3.1504, | |
| "step": 105350 | |
| }, | |
| { | |
| "epoch": 30.675652072659524, | |
| "grad_norm": 0.45064088702201843, | |
| "learning_rate": 0.00023203610949330223, | |
| "loss": 3.1418, | |
| "step": 105400 | |
| }, | |
| { | |
| "epoch": 30.69020726595249, | |
| "grad_norm": 0.48725900053977966, | |
| "learning_rate": 0.00023186138613861385, | |
| "loss": 3.1581, | |
| "step": 105450 | |
| }, | |
| { | |
| "epoch": 30.704762459245458, | |
| "grad_norm": 0.43696922063827515, | |
| "learning_rate": 0.00023168666278392542, | |
| "loss": 3.1458, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 30.719317652538425, | |
| "grad_norm": 0.45358753204345703, | |
| "learning_rate": 0.00023151193942923704, | |
| "loss": 3.1522, | |
| "step": 105550 | |
| }, | |
| { | |
| "epoch": 30.733872845831392, | |
| "grad_norm": 0.4805457592010498, | |
| "learning_rate": 0.0002313372160745486, | |
| "loss": 3.1392, | |
| "step": 105600 | |
| }, | |
| { | |
| "epoch": 30.74842803912436, | |
| "grad_norm": 0.44270044565200806, | |
| "learning_rate": 0.0002311624927198602, | |
| "loss": 3.1485, | |
| "step": 105650 | |
| }, | |
| { | |
| "epoch": 30.762983232417326, | |
| "grad_norm": 0.4392821192741394, | |
| "learning_rate": 0.0002309877693651718, | |
| "loss": 3.1522, | |
| "step": 105700 | |
| }, | |
| { | |
| "epoch": 30.777538425710294, | |
| "grad_norm": 0.47344040870666504, | |
| "learning_rate": 0.0002308130460104834, | |
| "loss": 3.1454, | |
| "step": 105750 | |
| }, | |
| { | |
| "epoch": 30.79209361900326, | |
| "grad_norm": 0.4679954946041107, | |
| "learning_rate": 0.00023063832265579496, | |
| "loss": 3.1484, | |
| "step": 105800 | |
| }, | |
| { | |
| "epoch": 30.806648812296228, | |
| "grad_norm": 0.4455646872520447, | |
| "learning_rate": 0.00023046359930110658, | |
| "loss": 3.1572, | |
| "step": 105850 | |
| }, | |
| { | |
| "epoch": 30.821204005589195, | |
| "grad_norm": 0.4564194679260254, | |
| "learning_rate": 0.00023028887594641814, | |
| "loss": 3.1502, | |
| "step": 105900 | |
| }, | |
| { | |
| "epoch": 30.83575919888216, | |
| "grad_norm": 0.4792768657207489, | |
| "learning_rate": 0.00023011415259172974, | |
| "loss": 3.1549, | |
| "step": 105950 | |
| }, | |
| { | |
| "epoch": 30.85031439217513, | |
| "grad_norm": 0.4297233819961548, | |
| "learning_rate": 0.00022993942923704133, | |
| "loss": 3.1584, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 30.85031439217513, | |
| "eval_accuracy": 0.3756258475922357, | |
| "eval_loss": 3.535869598388672, | |
| "eval_runtime": 179.0451, | |
| "eval_samples_per_second": 92.999, | |
| "eval_steps_per_second": 5.814, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 30.864869585468096, | |
| "grad_norm": 0.48069772124290466, | |
| "learning_rate": 0.00022976470588235293, | |
| "loss": 3.1533, | |
| "step": 106050 | |
| }, | |
| { | |
| "epoch": 30.879424778761063, | |
| "grad_norm": 0.4900089502334595, | |
| "learning_rate": 0.0002295899825276645, | |
| "loss": 3.1557, | |
| "step": 106100 | |
| }, | |
| { | |
| "epoch": 30.89397997205403, | |
| "grad_norm": 0.4526594281196594, | |
| "learning_rate": 0.00022941525917297612, | |
| "loss": 3.162, | |
| "step": 106150 | |
| }, | |
| { | |
| "epoch": 30.908535165346997, | |
| "grad_norm": 0.4509666860103607, | |
| "learning_rate": 0.00022924053581828768, | |
| "loss": 3.1507, | |
| "step": 106200 | |
| }, | |
| { | |
| "epoch": 30.923090358639964, | |
| "grad_norm": 0.44628453254699707, | |
| "learning_rate": 0.0002290658124635993, | |
| "loss": 3.1555, | |
| "step": 106250 | |
| }, | |
| { | |
| "epoch": 30.93764555193293, | |
| "grad_norm": 0.4314361810684204, | |
| "learning_rate": 0.00022889108910891087, | |
| "loss": 3.1643, | |
| "step": 106300 | |
| }, | |
| { | |
| "epoch": 30.9522007452259, | |
| "grad_norm": 0.44850438833236694, | |
| "learning_rate": 0.00022871636575422247, | |
| "loss": 3.1515, | |
| "step": 106350 | |
| }, | |
| { | |
| "epoch": 30.966755938518865, | |
| "grad_norm": 0.46688973903656006, | |
| "learning_rate": 0.00022854164239953406, | |
| "loss": 3.1586, | |
| "step": 106400 | |
| }, | |
| { | |
| "epoch": 30.98131113181183, | |
| "grad_norm": 0.46198490262031555, | |
| "learning_rate": 0.00022836691904484565, | |
| "loss": 3.1721, | |
| "step": 106450 | |
| }, | |
| { | |
| "epoch": 30.995866325104796, | |
| "grad_norm": 0.45322471857070923, | |
| "learning_rate": 0.00022819219569015722, | |
| "loss": 3.1583, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 31.010188635305077, | |
| "grad_norm": 0.46810445189476013, | |
| "learning_rate": 0.00022801747233546884, | |
| "loss": 3.1108, | |
| "step": 106550 | |
| }, | |
| { | |
| "epoch": 31.024743828598044, | |
| "grad_norm": 0.4760522246360779, | |
| "learning_rate": 0.0002278427489807804, | |
| "loss": 3.063, | |
| "step": 106600 | |
| }, | |
| { | |
| "epoch": 31.03929902189101, | |
| "grad_norm": 0.4931826889514923, | |
| "learning_rate": 0.00022766802562609198, | |
| "loss": 3.0837, | |
| "step": 106650 | |
| }, | |
| { | |
| "epoch": 31.05385421518398, | |
| "grad_norm": 0.456632137298584, | |
| "learning_rate": 0.0002274933022714036, | |
| "loss": 3.0864, | |
| "step": 106700 | |
| }, | |
| { | |
| "epoch": 31.068409408476946, | |
| "grad_norm": 0.45030152797698975, | |
| "learning_rate": 0.00022731857891671517, | |
| "loss": 3.0609, | |
| "step": 106750 | |
| }, | |
| { | |
| "epoch": 31.082964601769913, | |
| "grad_norm": 0.4781184792518616, | |
| "learning_rate": 0.00022714385556202676, | |
| "loss": 3.0877, | |
| "step": 106800 | |
| }, | |
| { | |
| "epoch": 31.09751979506288, | |
| "grad_norm": 0.4634675681591034, | |
| "learning_rate": 0.00022696913220733835, | |
| "loss": 3.0881, | |
| "step": 106850 | |
| }, | |
| { | |
| "epoch": 31.112074988355847, | |
| "grad_norm": 0.46249765157699585, | |
| "learning_rate": 0.00022679440885264995, | |
| "loss": 3.0788, | |
| "step": 106900 | |
| }, | |
| { | |
| "epoch": 31.126630181648814, | |
| "grad_norm": 0.46504703164100647, | |
| "learning_rate": 0.00022661968549796157, | |
| "loss": 3.0933, | |
| "step": 106950 | |
| }, | |
| { | |
| "epoch": 31.14118537494178, | |
| "grad_norm": 0.44973164796829224, | |
| "learning_rate": 0.00022644496214327314, | |
| "loss": 3.0948, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 31.14118537494178, | |
| "eval_accuracy": 0.37459031450424457, | |
| "eval_loss": 3.554295539855957, | |
| "eval_runtime": 181.1589, | |
| "eval_samples_per_second": 91.914, | |
| "eval_steps_per_second": 5.746, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 31.155740568234748, | |
| "grad_norm": 0.47105175256729126, | |
| "learning_rate": 0.0002262702387885847, | |
| "loss": 3.0937, | |
| "step": 107050 | |
| }, | |
| { | |
| "epoch": 31.17029576152771, | |
| "grad_norm": 0.4425552189350128, | |
| "learning_rate": 0.00022609551543389633, | |
| "loss": 3.0917, | |
| "step": 107100 | |
| }, | |
| { | |
| "epoch": 31.18485095482068, | |
| "grad_norm": 0.4712986350059509, | |
| "learning_rate": 0.0002259207920792079, | |
| "loss": 3.0927, | |
| "step": 107150 | |
| }, | |
| { | |
| "epoch": 31.199406148113646, | |
| "grad_norm": 0.4686832129955292, | |
| "learning_rate": 0.0002257460687245195, | |
| "loss": 3.0996, | |
| "step": 107200 | |
| }, | |
| { | |
| "epoch": 31.213961341406613, | |
| "grad_norm": 0.4482789933681488, | |
| "learning_rate": 0.00022557134536983108, | |
| "loss": 3.0915, | |
| "step": 107250 | |
| }, | |
| { | |
| "epoch": 31.22851653469958, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00022539662201514268, | |
| "loss": 3.1036, | |
| "step": 107300 | |
| }, | |
| { | |
| "epoch": 31.243071727992547, | |
| "grad_norm": 0.4736534059047699, | |
| "learning_rate": 0.00022522189866045424, | |
| "loss": 3.1067, | |
| "step": 107350 | |
| }, | |
| { | |
| "epoch": 31.257626921285514, | |
| "grad_norm": 0.44008541107177734, | |
| "learning_rate": 0.00022504717530576586, | |
| "loss": 3.0907, | |
| "step": 107400 | |
| }, | |
| { | |
| "epoch": 31.27218211457848, | |
| "grad_norm": 0.4680997133255005, | |
| "learning_rate": 0.00022487245195107743, | |
| "loss": 3.1064, | |
| "step": 107450 | |
| }, | |
| { | |
| "epoch": 31.286737307871448, | |
| "grad_norm": 0.4645274579524994, | |
| "learning_rate": 0.00022469772859638903, | |
| "loss": 3.111, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 31.301292501164415, | |
| "grad_norm": 0.43999406695365906, | |
| "learning_rate": 0.00022452300524170062, | |
| "loss": 3.1186, | |
| "step": 107550 | |
| }, | |
| { | |
| "epoch": 31.315847694457382, | |
| "grad_norm": 0.4434453547000885, | |
| "learning_rate": 0.00022434828188701221, | |
| "loss": 3.105, | |
| "step": 107600 | |
| }, | |
| { | |
| "epoch": 31.33040288775035, | |
| "grad_norm": 0.4525635242462158, | |
| "learning_rate": 0.0002241735585323238, | |
| "loss": 3.1177, | |
| "step": 107650 | |
| }, | |
| { | |
| "epoch": 31.344958081043316, | |
| "grad_norm": 0.4374854564666748, | |
| "learning_rate": 0.0002239988351776354, | |
| "loss": 3.1208, | |
| "step": 107700 | |
| }, | |
| { | |
| "epoch": 31.359513274336283, | |
| "grad_norm": 0.4741731584072113, | |
| "learning_rate": 0.00022382411182294697, | |
| "loss": 3.1195, | |
| "step": 107750 | |
| }, | |
| { | |
| "epoch": 31.37406846762925, | |
| "grad_norm": 0.4294305443763733, | |
| "learning_rate": 0.0002236493884682586, | |
| "loss": 3.115, | |
| "step": 107800 | |
| }, | |
| { | |
| "epoch": 31.388623660922217, | |
| "grad_norm": 0.44635891914367676, | |
| "learning_rate": 0.00022347466511357016, | |
| "loss": 3.1123, | |
| "step": 107850 | |
| }, | |
| { | |
| "epoch": 31.403178854215184, | |
| "grad_norm": 0.4601406455039978, | |
| "learning_rate": 0.00022329994175888175, | |
| "loss": 3.1098, | |
| "step": 107900 | |
| }, | |
| { | |
| "epoch": 31.41773404750815, | |
| "grad_norm": 0.4337484538555145, | |
| "learning_rate": 0.00022312521840419335, | |
| "loss": 3.1193, | |
| "step": 107950 | |
| }, | |
| { | |
| "epoch": 31.43228924080112, | |
| "grad_norm": 0.4300398826599121, | |
| "learning_rate": 0.00022295049504950494, | |
| "loss": 3.1236, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 31.43228924080112, | |
| "eval_accuracy": 0.3750808734770371, | |
| "eval_loss": 3.5484988689422607, | |
| "eval_runtime": 181.0232, | |
| "eval_samples_per_second": 91.983, | |
| "eval_steps_per_second": 5.751, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 31.446844434094086, | |
| "grad_norm": 0.5017353296279907, | |
| "learning_rate": 0.0002227757716948165, | |
| "loss": 3.126, | |
| "step": 108050 | |
| }, | |
| { | |
| "epoch": 31.461399627387053, | |
| "grad_norm": 0.45702579617500305, | |
| "learning_rate": 0.00022260104834012813, | |
| "loss": 3.1345, | |
| "step": 108100 | |
| }, | |
| { | |
| "epoch": 31.47595482068002, | |
| "grad_norm": 0.48082423210144043, | |
| "learning_rate": 0.0002224263249854397, | |
| "loss": 3.1295, | |
| "step": 108150 | |
| }, | |
| { | |
| "epoch": 31.490510013972987, | |
| "grad_norm": 0.4717453718185425, | |
| "learning_rate": 0.00022225160163075126, | |
| "loss": 3.1349, | |
| "step": 108200 | |
| }, | |
| { | |
| "epoch": 31.505065207265954, | |
| "grad_norm": 0.45219430327415466, | |
| "learning_rate": 0.00022207687827606289, | |
| "loss": 3.1237, | |
| "step": 108250 | |
| }, | |
| { | |
| "epoch": 31.51962040055892, | |
| "grad_norm": 0.46613046526908875, | |
| "learning_rate": 0.00022190215492137445, | |
| "loss": 3.1341, | |
| "step": 108300 | |
| }, | |
| { | |
| "epoch": 31.534175593851888, | |
| "grad_norm": 0.474324107170105, | |
| "learning_rate": 0.00022172743156668607, | |
| "loss": 3.125, | |
| "step": 108350 | |
| }, | |
| { | |
| "epoch": 31.548730787144855, | |
| "grad_norm": 0.47235554456710815, | |
| "learning_rate": 0.00022155270821199764, | |
| "loss": 3.1349, | |
| "step": 108400 | |
| }, | |
| { | |
| "epoch": 31.56328598043782, | |
| "grad_norm": 0.44595572352409363, | |
| "learning_rate": 0.00022137798485730924, | |
| "loss": 3.1368, | |
| "step": 108450 | |
| }, | |
| { | |
| "epoch": 31.577841173730786, | |
| "grad_norm": 0.4679282307624817, | |
| "learning_rate": 0.00022120326150262083, | |
| "loss": 3.1317, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 31.592396367023753, | |
| "grad_norm": 0.4556312561035156, | |
| "learning_rate": 0.00022102853814793242, | |
| "loss": 3.1197, | |
| "step": 108550 | |
| }, | |
| { | |
| "epoch": 31.60695156031672, | |
| "grad_norm": 0.47826388478279114, | |
| "learning_rate": 0.000220853814793244, | |
| "loss": 3.1234, | |
| "step": 108600 | |
| }, | |
| { | |
| "epoch": 31.621506753609687, | |
| "grad_norm": 0.48391714692115784, | |
| "learning_rate": 0.0002206790914385556, | |
| "loss": 3.1286, | |
| "step": 108650 | |
| }, | |
| { | |
| "epoch": 31.636061946902654, | |
| "grad_norm": 0.47049498558044434, | |
| "learning_rate": 0.00022050436808386718, | |
| "loss": 3.1443, | |
| "step": 108700 | |
| }, | |
| { | |
| "epoch": 31.65061714019562, | |
| "grad_norm": 0.45760464668273926, | |
| "learning_rate": 0.00022032964472917877, | |
| "loss": 3.1455, | |
| "step": 108750 | |
| }, | |
| { | |
| "epoch": 31.665172333488588, | |
| "grad_norm": 0.4569242298603058, | |
| "learning_rate": 0.00022015492137449037, | |
| "loss": 3.1406, | |
| "step": 108800 | |
| }, | |
| { | |
| "epoch": 31.679727526781555, | |
| "grad_norm": 0.4866321384906769, | |
| "learning_rate": 0.00021998019801980196, | |
| "loss": 3.1426, | |
| "step": 108850 | |
| }, | |
| { | |
| "epoch": 31.694282720074522, | |
| "grad_norm": 0.4808866083621979, | |
| "learning_rate": 0.00021980547466511353, | |
| "loss": 3.1362, | |
| "step": 108900 | |
| }, | |
| { | |
| "epoch": 31.70883791336749, | |
| "grad_norm": 0.44989365339279175, | |
| "learning_rate": 0.00021963075131042515, | |
| "loss": 3.1458, | |
| "step": 108950 | |
| }, | |
| { | |
| "epoch": 31.723393106660456, | |
| "grad_norm": 0.456107497215271, | |
| "learning_rate": 0.00021945602795573672, | |
| "loss": 3.1344, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 31.723393106660456, | |
| "eval_accuracy": 0.3755069099591581, | |
| "eval_loss": 3.5397472381591797, | |
| "eval_runtime": 180.9711, | |
| "eval_samples_per_second": 92.009, | |
| "eval_steps_per_second": 5.752, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 31.737948299953423, | |
| "grad_norm": 0.44981271028518677, | |
| "learning_rate": 0.00021928130460104834, | |
| "loss": 3.1351, | |
| "step": 109050 | |
| }, | |
| { | |
| "epoch": 31.75250349324639, | |
| "grad_norm": 0.48247814178466797, | |
| "learning_rate": 0.0002191065812463599, | |
| "loss": 3.1416, | |
| "step": 109100 | |
| }, | |
| { | |
| "epoch": 31.767058686539357, | |
| "grad_norm": 0.4782215356826782, | |
| "learning_rate": 0.0002189318578916715, | |
| "loss": 3.1548, | |
| "step": 109150 | |
| }, | |
| { | |
| "epoch": 31.781613879832324, | |
| "grad_norm": 0.4576863944530487, | |
| "learning_rate": 0.0002187571345369831, | |
| "loss": 3.153, | |
| "step": 109200 | |
| }, | |
| { | |
| "epoch": 31.79616907312529, | |
| "grad_norm": 0.46203556656837463, | |
| "learning_rate": 0.0002185824111822947, | |
| "loss": 3.1496, | |
| "step": 109250 | |
| }, | |
| { | |
| "epoch": 31.81072426641826, | |
| "grad_norm": 0.45210832357406616, | |
| "learning_rate": 0.00021840768782760626, | |
| "loss": 3.1359, | |
| "step": 109300 | |
| }, | |
| { | |
| "epoch": 31.825279459711226, | |
| "grad_norm": 0.4440477788448334, | |
| "learning_rate": 0.00021823296447291788, | |
| "loss": 3.139, | |
| "step": 109350 | |
| }, | |
| { | |
| "epoch": 31.839834653004193, | |
| "grad_norm": 0.47411999106407166, | |
| "learning_rate": 0.00021805824111822945, | |
| "loss": 3.1456, | |
| "step": 109400 | |
| }, | |
| { | |
| "epoch": 31.85438984629716, | |
| "grad_norm": 0.46934619545936584, | |
| "learning_rate": 0.00021788351776354104, | |
| "loss": 3.1523, | |
| "step": 109450 | |
| }, | |
| { | |
| "epoch": 31.868945039590127, | |
| "grad_norm": 0.4774060547351837, | |
| "learning_rate": 0.00021770879440885263, | |
| "loss": 3.1483, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 31.883500232883094, | |
| "grad_norm": 0.4577077329158783, | |
| "learning_rate": 0.00021753407105416423, | |
| "loss": 3.1504, | |
| "step": 109550 | |
| }, | |
| { | |
| "epoch": 31.89805542617606, | |
| "grad_norm": 0.4703179597854614, | |
| "learning_rate": 0.0002173593476994758, | |
| "loss": 3.1481, | |
| "step": 109600 | |
| }, | |
| { | |
| "epoch": 31.912610619469028, | |
| "grad_norm": 0.4769929349422455, | |
| "learning_rate": 0.00021718462434478742, | |
| "loss": 3.1508, | |
| "step": 109650 | |
| }, | |
| { | |
| "epoch": 31.927165812761995, | |
| "grad_norm": 0.49723508954048157, | |
| "learning_rate": 0.00021700990099009898, | |
| "loss": 3.156, | |
| "step": 109700 | |
| }, | |
| { | |
| "epoch": 31.941721006054962, | |
| "grad_norm": 0.4571239650249481, | |
| "learning_rate": 0.0002168351776354106, | |
| "loss": 3.1584, | |
| "step": 109750 | |
| }, | |
| { | |
| "epoch": 31.956276199347926, | |
| "grad_norm": 0.4813500642776489, | |
| "learning_rate": 0.00021666045428072217, | |
| "loss": 3.138, | |
| "step": 109800 | |
| }, | |
| { | |
| "epoch": 31.970831392640893, | |
| "grad_norm": 0.4950752258300781, | |
| "learning_rate": 0.00021648573092603374, | |
| "loss": 3.1532, | |
| "step": 109850 | |
| }, | |
| { | |
| "epoch": 31.98538658593386, | |
| "grad_norm": 0.4535057842731476, | |
| "learning_rate": 0.00021631100757134536, | |
| "loss": 3.1451, | |
| "step": 109900 | |
| }, | |
| { | |
| "epoch": 31.999941779226827, | |
| "grad_norm": 0.46446654200553894, | |
| "learning_rate": 0.00021613628421665693, | |
| "loss": 3.1475, | |
| "step": 109950 | |
| }, | |
| { | |
| "epoch": 32.01426408942711, | |
| "grad_norm": 0.4385596513748169, | |
| "learning_rate": 0.00021596156086196852, | |
| "loss": 3.0595, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 32.01426408942711, | |
| "eval_accuracy": 0.37491116404802116, | |
| "eval_loss": 3.550321340560913, | |
| "eval_runtime": 179.2352, | |
| "eval_samples_per_second": 92.9, | |
| "eval_steps_per_second": 5.808, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 32.01426408942711, | |
| "step": 110000, | |
| "total_flos": 2.298801310138368e+18, | |
| "train_loss": 0.5713334234064276, | |
| "train_runtime": 39943.0578, | |
| "train_samples_per_second": 343.998, | |
| "train_steps_per_second": 4.301 | |
| } | |
| ], | |
| "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": 20 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.298801310138368e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |