Invalid JSON: Unexpected token 'N', ..."ad_norm": NaN,
"... is not valid JSON
| { | |
| "best_global_step": 79000, | |
| "best_metric": 3.52482271194458, | |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/swap_require_to_hit_3591/checkpoint-30000", | |
| "epoch": 28.812761993479274, | |
| "eval_steps": 1000, | |
| "global_step": 99000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014555193292966931, | |
| "grad_norm": 1.174659013748169, | |
| "learning_rate": 0.000294, | |
| "loss": 8.4389, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029110386585933862, | |
| "grad_norm": 0.9201799035072327, | |
| "learning_rate": 0.0005939999999999999, | |
| "loss": 6.7398, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04366557987890079, | |
| "grad_norm": 0.5460855960845947, | |
| "learning_rate": 0.0005998287711124053, | |
| "loss": 6.3813, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.058220773171867725, | |
| "grad_norm": 0.44160327315330505, | |
| "learning_rate": 0.000599654047757717, | |
| "loss": 6.1597, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07277596646483465, | |
| "grad_norm": 0.5325834155082703, | |
| "learning_rate": 0.0005994793244030285, | |
| "loss": 6.0122, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.08733115975780158, | |
| "grad_norm": 0.4245791435241699, | |
| "learning_rate": 0.00059930460104834, | |
| "loss": 5.8907, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.10188635305076851, | |
| "grad_norm": 0.4183367192745209, | |
| "learning_rate": 0.0005991298776936517, | |
| "loss": 5.7558, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.11644154634373545, | |
| "grad_norm": 0.4702097475528717, | |
| "learning_rate": 0.0005989551543389632, | |
| "loss": 5.6455, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1309967396367024, | |
| "grad_norm": 0.4629879891872406, | |
| "learning_rate": 0.0005987804309842748, | |
| "loss": 5.5292, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1455519329296693, | |
| "grad_norm": 0.41067075729370117, | |
| "learning_rate": 0.0005986057076295864, | |
| "loss": 5.4179, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16010712622263623, | |
| "grad_norm": 0.45282429456710815, | |
| "learning_rate": 0.0005984309842748981, | |
| "loss": 5.3436, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.17466231951560315, | |
| "grad_norm": 0.45462384819984436, | |
| "learning_rate": 0.0005982562609202096, | |
| "loss": 5.2624, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1892175128085701, | |
| "grad_norm": 0.5746437311172485, | |
| "learning_rate": 0.0005980815375655212, | |
| "loss": 5.1987, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.20377270610153703, | |
| "grad_norm": 0.416387140750885, | |
| "learning_rate": 0.0005979068142108328, | |
| "loss": 5.1288, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.21832789939450395, | |
| "grad_norm": 0.4375811219215393, | |
| "learning_rate": 0.0005977320908561445, | |
| "loss": 5.0761, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2328830926874709, | |
| "grad_norm": 0.40442579984664917, | |
| "learning_rate": 0.000597557367501456, | |
| "loss": 5.0175, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.24743828598043782, | |
| "grad_norm": 0.48197323083877563, | |
| "learning_rate": 0.0005973826441467675, | |
| "loss": 4.9754, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2619934792734048, | |
| "grad_norm": 0.4318915903568268, | |
| "learning_rate": 0.0005972079207920792, | |
| "loss": 4.9245, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.27654867256637167, | |
| "grad_norm": 0.45740172266960144, | |
| "learning_rate": 0.0005970331974373907, | |
| "loss": 4.8692, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "grad_norm": 0.4700559973716736, | |
| "learning_rate": 0.0005968584740827023, | |
| "loss": 4.8412, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2911038658593386, | |
| "eval_accuracy": 0.2532508934131939, | |
| "eval_loss": 4.7631516456604, | |
| "eval_runtime": 83.8317, | |
| "eval_samples_per_second": 198.624, | |
| "eval_steps_per_second": 12.418, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.30565905915230557, | |
| "grad_norm": 0.4623873829841614, | |
| "learning_rate": 0.0005966837507280139, | |
| "loss": 4.7849, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.32021425244527246, | |
| "grad_norm": 0.4560273587703705, | |
| "learning_rate": 0.0005965090273733256, | |
| "loss": 4.7479, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3347694457382394, | |
| "grad_norm": 0.4368627071380615, | |
| "learning_rate": 0.0005963343040186371, | |
| "loss": 4.7059, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3493246390312063, | |
| "grad_norm": 0.431756854057312, | |
| "learning_rate": 0.0005961595806639486, | |
| "loss": 4.664, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.36387983232417326, | |
| "grad_norm": 0.45245376229286194, | |
| "learning_rate": 0.0005959848573092603, | |
| "loss": 4.6251, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.3784350256171402, | |
| "grad_norm": 0.5368194580078125, | |
| "learning_rate": 0.0005958101339545718, | |
| "loss": 4.6054, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3929902189101071, | |
| "grad_norm": 0.45866209268569946, | |
| "learning_rate": 0.0005956354105998835, | |
| "loss": 4.5729, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.40754541220307405, | |
| "grad_norm": 0.46009761095046997, | |
| "learning_rate": 0.000595460687245195, | |
| "loss": 4.5446, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.422100605496041, | |
| "grad_norm": 0.4195514917373657, | |
| "learning_rate": 0.0005952859638905067, | |
| "loss": 4.5334, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.4366557987890079, | |
| "grad_norm": 0.3959258198738098, | |
| "learning_rate": 0.0005951112405358182, | |
| "loss": 4.511, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.45121099208197485, | |
| "grad_norm": 0.45320427417755127, | |
| "learning_rate": 0.0005949365171811299, | |
| "loss": 4.4808, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4657661853749418, | |
| "grad_norm": 0.44871363043785095, | |
| "learning_rate": 0.0005947617938264414, | |
| "loss": 4.4679, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.4803213786679087, | |
| "grad_norm": 0.40834811329841614, | |
| "learning_rate": 0.000594587070471753, | |
| "loss": 4.4538, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.49487657196087564, | |
| "grad_norm": 0.39514100551605225, | |
| "learning_rate": 0.0005944123471170646, | |
| "loss": 4.4261, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5094317652538426, | |
| "grad_norm": 0.3988879919052124, | |
| "learning_rate": 0.0005942376237623762, | |
| "loss": 4.4111, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5239869585468095, | |
| "grad_norm": 0.41054731607437134, | |
| "learning_rate": 0.0005940629004076878, | |
| "loss": 4.3975, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5385421518397764, | |
| "grad_norm": 0.4006456136703491, | |
| "learning_rate": 0.0005938881770529993, | |
| "loss": 4.3816, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5530973451327433, | |
| "grad_norm": 0.39680641889572144, | |
| "learning_rate": 0.000593713453698311, | |
| "loss": 4.3709, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.5676525384257103, | |
| "grad_norm": 0.39294660091400146, | |
| "learning_rate": 0.0005935387303436226, | |
| "loss": 4.3468, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "grad_norm": 0.4436911940574646, | |
| "learning_rate": 0.0005933640069889342, | |
| "loss": 4.3374, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "eval_accuracy": 0.29963468987658576, | |
| "eval_loss": 4.27892541885376, | |
| "eval_runtime": 83.7147, | |
| "eval_samples_per_second": 198.902, | |
| "eval_steps_per_second": 12.435, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5967629250116442, | |
| "grad_norm": 0.4191906750202179, | |
| "learning_rate": 0.0005931892836342457, | |
| "loss": 4.3228, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.6113181183046111, | |
| "grad_norm": 0.4014892280101776, | |
| "learning_rate": 0.0005930145602795573, | |
| "loss": 4.3074, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.625873311597578, | |
| "grad_norm": 0.403000146150589, | |
| "learning_rate": 0.000592839836924869, | |
| "loss": 4.297, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6404285048905449, | |
| "grad_norm": 0.40409624576568604, | |
| "learning_rate": 0.0005926651135701805, | |
| "loss": 4.2981, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.6549836981835119, | |
| "grad_norm": 0.3735470771789551, | |
| "learning_rate": 0.0005924903902154921, | |
| "loss": 4.2796, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6695388914764788, | |
| "grad_norm": 0.40209901332855225, | |
| "learning_rate": 0.0005923156668608037, | |
| "loss": 4.2629, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.6840940847694458, | |
| "grad_norm": 0.3694210648536682, | |
| "learning_rate": 0.0005921409435061153, | |
| "loss": 4.2575, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6986492780624126, | |
| "grad_norm": 0.4106614291667938, | |
| "learning_rate": 0.0005919662201514268, | |
| "loss": 4.2362, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7132044713553796, | |
| "grad_norm": 0.3787236511707306, | |
| "learning_rate": 0.0005917914967967384, | |
| "loss": 4.2403, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.7277596646483465, | |
| "grad_norm": 0.4535953104496002, | |
| "learning_rate": 0.0005916167734420501, | |
| "loss": 4.217, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7423148579413135, | |
| "grad_norm": 0.3803585469722748, | |
| "learning_rate": 0.0005914420500873616, | |
| "loss": 4.2206, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7568700512342804, | |
| "grad_norm": 0.39158543944358826, | |
| "learning_rate": 0.0005912673267326732, | |
| "loss": 4.2058, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.7714252445272474, | |
| "grad_norm": 0.3906506597995758, | |
| "learning_rate": 0.0005910926033779848, | |
| "loss": 4.2003, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.7859804378202142, | |
| "grad_norm": 0.38841021060943604, | |
| "learning_rate": 0.0005909178800232964, | |
| "loss": 4.1806, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8005356311131812, | |
| "grad_norm": 0.36479631066322327, | |
| "learning_rate": 0.000590743156668608, | |
| "loss": 4.1831, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.8150908244061481, | |
| "grad_norm": 0.3638639450073242, | |
| "learning_rate": 0.0005905684333139196, | |
| "loss": 4.1747, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8296460176991151, | |
| "grad_norm": 0.39287617802619934, | |
| "learning_rate": 0.0005903937099592312, | |
| "loss": 4.1641, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.844201210992082, | |
| "grad_norm": 0.34700122475624084, | |
| "learning_rate": 0.0005902189866045427, | |
| "loss": 4.1552, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.858756404285049, | |
| "grad_norm": 0.36135420203208923, | |
| "learning_rate": 0.0005900442632498543, | |
| "loss": 4.1415, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "grad_norm": 0.37881818413734436, | |
| "learning_rate": 0.0005898695398951659, | |
| "loss": 4.1421, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8733115975780158, | |
| "eval_accuracy": 0.31618218189677555, | |
| "eval_loss": 4.088852405548096, | |
| "eval_runtime": 83.6984, | |
| "eval_samples_per_second": 198.94, | |
| "eval_steps_per_second": 12.438, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8878667908709827, | |
| "grad_norm": 0.35876956582069397, | |
| "learning_rate": 0.0005896948165404776, | |
| "loss": 4.1416, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.9024219841639497, | |
| "grad_norm": 0.3281998336315155, | |
| "learning_rate": 0.0005895200931857891, | |
| "loss": 4.1235, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9169771774569166, | |
| "grad_norm": 0.34561455249786377, | |
| "learning_rate": 0.0005893453698311007, | |
| "loss": 4.1292, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9315323707498836, | |
| "grad_norm": 0.3581830561161041, | |
| "learning_rate": 0.0005891706464764123, | |
| "loss": 4.1061, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.9460875640428504, | |
| "grad_norm": 0.346056193113327, | |
| "learning_rate": 0.0005889959231217238, | |
| "loss": 4.1105, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.9606427573358174, | |
| "grad_norm": 0.3447182774543762, | |
| "learning_rate": 0.0005888211997670355, | |
| "loss": 4.1084, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.9751979506287843, | |
| "grad_norm": 0.3601776957511902, | |
| "learning_rate": 0.000588646476412347, | |
| "loss": 4.1033, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.9897531439217513, | |
| "grad_norm": 0.3602287173271179, | |
| "learning_rate": 0.0005884717530576587, | |
| "loss": 4.103, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.0040754541220307, | |
| "grad_norm": 0.33607354760169983, | |
| "learning_rate": 0.0005882970297029702, | |
| "loss": 4.0694, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.0186306474149978, | |
| "grad_norm": 0.32804882526397705, | |
| "learning_rate": 0.0005881223063482818, | |
| "loss": 4.0145, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.0331858407079646, | |
| "grad_norm": 0.3770780861377716, | |
| "learning_rate": 0.0005879475829935934, | |
| "loss": 4.0215, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.0477410340009314, | |
| "grad_norm": 0.3616582751274109, | |
| "learning_rate": 0.0005877728596389051, | |
| "loss": 4.0122, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.0622962272938985, | |
| "grad_norm": 0.3846186101436615, | |
| "learning_rate": 0.0005875981362842166, | |
| "loss": 4.0107, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.0768514205868653, | |
| "grad_norm": 0.39962780475616455, | |
| "learning_rate": 0.0005874234129295281, | |
| "loss": 4.0294, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.0914066138798324, | |
| "grad_norm": 0.33880868554115295, | |
| "learning_rate": 0.0005872486895748398, | |
| "loss": 4.0056, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.1059618071727992, | |
| "grad_norm": 0.3509419858455658, | |
| "learning_rate": 0.0005870739662201513, | |
| "loss": 4.0025, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.120517000465766, | |
| "grad_norm": 0.3512333929538727, | |
| "learning_rate": 0.000586899242865463, | |
| "loss": 3.9953, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.1350721937587331, | |
| "grad_norm": 0.3627850115299225, | |
| "learning_rate": 0.0005867245195107746, | |
| "loss": 4.0, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.1496273870517, | |
| "grad_norm": 0.3500107228755951, | |
| "learning_rate": 0.0005865497961560862, | |
| "loss": 4.0007, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "grad_norm": 0.36411580443382263, | |
| "learning_rate": 0.0005863750728013977, | |
| "loss": 3.9896, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.164182580344667, | |
| "eval_accuracy": 0.32585973280637226, | |
| "eval_loss": 3.9891650676727295, | |
| "eval_runtime": 83.6716, | |
| "eval_samples_per_second": 199.004, | |
| "eval_steps_per_second": 12.441, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.1787377736376339, | |
| "grad_norm": 0.3315863311290741, | |
| "learning_rate": 0.0005862003494467094, | |
| "loss": 3.9779, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.193292966930601, | |
| "grad_norm": 0.3490806818008423, | |
| "learning_rate": 0.0005860256260920209, | |
| "loss": 3.9981, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.2078481602235678, | |
| "grad_norm": 0.35695990920066833, | |
| "learning_rate": 0.0005858509027373325, | |
| "loss": 3.9902, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 1.2224033535165346, | |
| "grad_norm": 0.3498048484325409, | |
| "learning_rate": 0.0005856761793826441, | |
| "loss": 3.9641, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.2369585468095017, | |
| "grad_norm": 0.34359222650527954, | |
| "learning_rate": 0.0005855014560279557, | |
| "loss": 3.9725, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 1.2515137401024685, | |
| "grad_norm": 0.32852908968925476, | |
| "learning_rate": 0.0005853267326732673, | |
| "loss": 3.9625, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.2660689333954354, | |
| "grad_norm": 0.3527366816997528, | |
| "learning_rate": 0.0005851520093185788, | |
| "loss": 3.9548, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 1.2806241266884024, | |
| "grad_norm": 0.3435719311237335, | |
| "learning_rate": 0.0005849772859638905, | |
| "loss": 3.953, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.2951793199813695, | |
| "grad_norm": 0.33108583092689514, | |
| "learning_rate": 0.0005848025626092021, | |
| "loss": 3.9681, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 1.3097345132743363, | |
| "grad_norm": 0.34129810333251953, | |
| "learning_rate": 0.0005846278392545136, | |
| "loss": 3.9564, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.3242897065673032, | |
| "grad_norm": 0.3452935218811035, | |
| "learning_rate": 0.0005844531158998252, | |
| "loss": 3.9473, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.3388448998602702, | |
| "grad_norm": 0.3214864730834961, | |
| "learning_rate": 0.0005842783925451368, | |
| "loss": 3.939, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.353400093153237, | |
| "grad_norm": 0.3587154150009155, | |
| "learning_rate": 0.0005841036691904484, | |
| "loss": 3.9414, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.367955286446204, | |
| "grad_norm": 0.3252437114715576, | |
| "learning_rate": 0.00058392894583576, | |
| "loss": 3.9439, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.382510479739171, | |
| "grad_norm": 0.3494246006011963, | |
| "learning_rate": 0.0005837542224810716, | |
| "loss": 3.9409, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 1.3970656730321378, | |
| "grad_norm": 0.3640254735946655, | |
| "learning_rate": 0.0005835794991263832, | |
| "loss": 3.9382, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.4116208663251049, | |
| "grad_norm": 0.3287889361381531, | |
| "learning_rate": 0.0005834047757716948, | |
| "loss": 3.9312, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 1.4261760596180717, | |
| "grad_norm": 0.3492181897163391, | |
| "learning_rate": 0.0005832300524170063, | |
| "loss": 3.9376, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.4407312529110388, | |
| "grad_norm": 0.35362449288368225, | |
| "learning_rate": 0.0005830553290623179, | |
| "loss": 3.9299, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "grad_norm": 0.3251268267631531, | |
| "learning_rate": 0.0005828806057076296, | |
| "loss": 3.9315, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4552864462040056, | |
| "eval_accuracy": 0.33212205774798176, | |
| "eval_loss": 3.911984920501709, | |
| "eval_runtime": 83.8514, | |
| "eval_samples_per_second": 198.577, | |
| "eval_steps_per_second": 12.415, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.4698416394969724, | |
| "grad_norm": 0.360942006111145, | |
| "learning_rate": 0.0005827058823529411, | |
| "loss": 3.92, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 1.4843968327899395, | |
| "grad_norm": 0.35775816440582275, | |
| "learning_rate": 0.0005825311589982527, | |
| "loss": 3.9265, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.4989520260829063, | |
| "grad_norm": 0.32437118887901306, | |
| "learning_rate": 0.0005823564356435643, | |
| "loss": 3.912, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 1.5135072193758732, | |
| "grad_norm": 0.3212335407733917, | |
| "learning_rate": 0.0005821817122888759, | |
| "loss": 3.9298, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.5280624126688402, | |
| "grad_norm": 0.34615692496299744, | |
| "learning_rate": 0.0005820069889341875, | |
| "loss": 3.9256, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 1.5426176059618073, | |
| "grad_norm": 0.34024184942245483, | |
| "learning_rate": 0.000581832265579499, | |
| "loss": 3.9066, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.5571727992547741, | |
| "grad_norm": 0.333649218082428, | |
| "learning_rate": 0.0005816575422248107, | |
| "loss": 3.9184, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 1.571727992547741, | |
| "grad_norm": 0.3335813283920288, | |
| "learning_rate": 0.0005814828188701222, | |
| "loss": 3.905, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.586283185840708, | |
| "grad_norm": 0.31955406069755554, | |
| "learning_rate": 0.0005813080955154338, | |
| "loss": 3.914, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.6008383791336749, | |
| "grad_norm": 0.33208370208740234, | |
| "learning_rate": 0.0005811333721607454, | |
| "loss": 3.8923, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.6153935724266417, | |
| "grad_norm": 0.32438549399375916, | |
| "learning_rate": 0.0005809586488060571, | |
| "loss": 3.8882, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.6299487657196088, | |
| "grad_norm": 0.3385221064090729, | |
| "learning_rate": 0.0005807839254513686, | |
| "loss": 3.8921, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.6445039590125758, | |
| "grad_norm": 0.33085668087005615, | |
| "learning_rate": 0.0005806092020966802, | |
| "loss": 3.8902, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.6590591523055425, | |
| "grad_norm": 0.3079417645931244, | |
| "learning_rate": 0.0005804344787419918, | |
| "loss": 3.8878, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.6736143455985095, | |
| "grad_norm": 0.33513057231903076, | |
| "learning_rate": 0.0005802597553873033, | |
| "loss": 3.8839, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.6881695388914766, | |
| "grad_norm": 0.34096869826316833, | |
| "learning_rate": 0.000580085032032615, | |
| "loss": 3.8812, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.7027247321844434, | |
| "grad_norm": 0.33812931180000305, | |
| "learning_rate": 0.0005799103086779265, | |
| "loss": 3.8819, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.7172799254774103, | |
| "grad_norm": 0.31069278717041016, | |
| "learning_rate": 0.0005797355853232382, | |
| "loss": 3.8668, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.7318351187703773, | |
| "grad_norm": 0.3248761296272278, | |
| "learning_rate": 0.0005795608619685497, | |
| "loss": 3.8825, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "grad_norm": 0.3348195254802704, | |
| "learning_rate": 0.0005793861386138614, | |
| "loss": 3.8626, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.7463903120633442, | |
| "eval_accuracy": 0.3376635877254952, | |
| "eval_loss": 3.853886365890503, | |
| "eval_runtime": 83.7816, | |
| "eval_samples_per_second": 198.743, | |
| "eval_steps_per_second": 12.425, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.760945505356311, | |
| "grad_norm": 0.3087477385997772, | |
| "learning_rate": 0.0005792114152591729, | |
| "loss": 3.8752, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.775500698649278, | |
| "grad_norm": 0.34609517455101013, | |
| "learning_rate": 0.0005790366919044846, | |
| "loss": 3.8843, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.7900558919422451, | |
| "grad_norm": 0.3276951313018799, | |
| "learning_rate": 0.0005788619685497961, | |
| "loss": 3.867, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.804611085235212, | |
| "grad_norm": 0.33969834446907043, | |
| "learning_rate": 0.0005786872451951077, | |
| "loss": 3.857, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.8191662785281788, | |
| "grad_norm": 0.3212492763996124, | |
| "learning_rate": 0.0005785125218404193, | |
| "loss": 3.8653, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.8337214718211459, | |
| "grad_norm": 0.3264020085334778, | |
| "learning_rate": 0.0005783377984857308, | |
| "loss": 3.8714, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.8482766651141127, | |
| "grad_norm": 0.32655003666877747, | |
| "learning_rate": 0.0005781630751310425, | |
| "loss": 3.8676, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.8628318584070795, | |
| "grad_norm": 0.31611907482147217, | |
| "learning_rate": 0.0005779883517763541, | |
| "loss": 3.8598, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.8773870517000466, | |
| "grad_norm": 0.34274324774742126, | |
| "learning_rate": 0.0005778136284216657, | |
| "loss": 3.8501, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.8919422449930137, | |
| "grad_norm": 0.3349476456642151, | |
| "learning_rate": 0.0005776389050669772, | |
| "loss": 3.8617, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.9064974382859803, | |
| "grad_norm": 0.33388906717300415, | |
| "learning_rate": 0.0005774641817122889, | |
| "loss": 3.8384, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.9210526315789473, | |
| "grad_norm": 0.3323456645011902, | |
| "learning_rate": 0.0005772894583576004, | |
| "loss": 3.8585, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.9356078248719144, | |
| "grad_norm": 0.3317473232746124, | |
| "learning_rate": 0.000577114735002912, | |
| "loss": 3.8452, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.9501630181648812, | |
| "grad_norm": 0.31404414772987366, | |
| "learning_rate": 0.0005769400116482236, | |
| "loss": 3.8376, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.964718211457848, | |
| "grad_norm": 0.32566794753074646, | |
| "learning_rate": 0.0005767652882935352, | |
| "loss": 3.8541, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.9792734047508151, | |
| "grad_norm": 0.34173375368118286, | |
| "learning_rate": 0.0005765905649388468, | |
| "loss": 3.8489, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.993828598043782, | |
| "grad_norm": 0.3266793489456177, | |
| "learning_rate": 0.0005764158415841583, | |
| "loss": 3.8364, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 2.0081509082440614, | |
| "grad_norm": 0.3377951979637146, | |
| "learning_rate": 0.00057624111822947, | |
| "loss": 3.7798, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.0227061015370285, | |
| "grad_norm": 0.3201465904712677, | |
| "learning_rate": 0.0005760663948747816, | |
| "loss": 3.731, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "grad_norm": 0.31822001934051514, | |
| "learning_rate": 0.0005758916715200931, | |
| "loss": 3.7296, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.0372612948299955, | |
| "eval_accuracy": 0.34207568029799285, | |
| "eval_loss": 3.8099608421325684, | |
| "eval_runtime": 83.7546, | |
| "eval_samples_per_second": 198.807, | |
| "eval_steps_per_second": 12.429, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.051816488122962, | |
| "grad_norm": 0.32309970259666443, | |
| "learning_rate": 0.0005757169481654047, | |
| "loss": 3.7427, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 2.066371681415929, | |
| "grad_norm": 0.32308658957481384, | |
| "learning_rate": 0.0005755422248107163, | |
| "loss": 3.7433, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.0809268747088963, | |
| "grad_norm": 0.3399259150028229, | |
| "learning_rate": 0.0005753675014560279, | |
| "loss": 3.7494, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 2.095482068001863, | |
| "grad_norm": 0.3357674479484558, | |
| "learning_rate": 0.0005751927781013395, | |
| "loss": 3.7456, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.11003726129483, | |
| "grad_norm": 0.3079433739185333, | |
| "learning_rate": 0.0005750180547466511, | |
| "loss": 3.7625, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 2.124592454587797, | |
| "grad_norm": 0.3221086859703064, | |
| "learning_rate": 0.0005748433313919627, | |
| "loss": 3.7508, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 2.139147647880764, | |
| "grad_norm": 0.3263172507286072, | |
| "learning_rate": 0.0005746686080372743, | |
| "loss": 3.7467, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 2.1537028411737307, | |
| "grad_norm": 0.3229222893714905, | |
| "learning_rate": 0.0005744938846825858, | |
| "loss": 3.7392, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.1682580344666977, | |
| "grad_norm": 0.31585898995399475, | |
| "learning_rate": 0.0005743191613278974, | |
| "loss": 3.7587, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 2.182813227759665, | |
| "grad_norm": 0.3362402617931366, | |
| "learning_rate": 0.0005741444379732091, | |
| "loss": 3.7556, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.1973684210526314, | |
| "grad_norm": 0.338635116815567, | |
| "learning_rate": 0.0005739697146185206, | |
| "loss": 3.753, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 2.2119236143455985, | |
| "grad_norm": 0.3221960663795471, | |
| "learning_rate": 0.0005737949912638322, | |
| "loss": 3.7637, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.2264788076385655, | |
| "grad_norm": 0.3351588845252991, | |
| "learning_rate": 0.0005736202679091438, | |
| "loss": 3.7483, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 2.241034000931532, | |
| "grad_norm": 0.32203781604766846, | |
| "learning_rate": 0.0005734455445544554, | |
| "loss": 3.7542, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 2.255589194224499, | |
| "grad_norm": 0.30380573868751526, | |
| "learning_rate": 0.000573270821199767, | |
| "loss": 3.7378, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 2.2701443875174663, | |
| "grad_norm": 0.3412049412727356, | |
| "learning_rate": 0.0005730960978450785, | |
| "loss": 3.7534, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.2846995808104333, | |
| "grad_norm": 0.3314678370952606, | |
| "learning_rate": 0.0005729213744903902, | |
| "loss": 3.7477, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 2.2992547741034, | |
| "grad_norm": 0.30382969975471497, | |
| "learning_rate": 0.0005727466511357017, | |
| "loss": 3.7625, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 2.313809967396367, | |
| "grad_norm": 0.3188897669315338, | |
| "learning_rate": 0.0005725719277810134, | |
| "loss": 3.7536, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "grad_norm": 0.30651071667671204, | |
| "learning_rate": 0.0005723972044263249, | |
| "loss": 3.7404, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.328365160689334, | |
| "eval_accuracy": 0.3450303167560677, | |
| "eval_loss": 3.7775912284851074, | |
| "eval_runtime": 83.6745, | |
| "eval_samples_per_second": 198.997, | |
| "eval_steps_per_second": 12.441, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.3429203539823007, | |
| "grad_norm": 0.308830589056015, | |
| "learning_rate": 0.0005722224810716366, | |
| "loss": 3.7682, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 2.3574755472752678, | |
| "grad_norm": 0.31910064816474915, | |
| "learning_rate": 0.0005720477577169481, | |
| "loss": 3.7482, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 2.372030740568235, | |
| "grad_norm": 0.32413697242736816, | |
| "learning_rate": 0.0005718730343622598, | |
| "loss": 3.7498, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 2.386585933861202, | |
| "grad_norm": 0.3139588534832001, | |
| "learning_rate": 0.0005716983110075713, | |
| "loss": 3.7398, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.4011411271541685, | |
| "grad_norm": 0.30236372351646423, | |
| "learning_rate": 0.0005715235876528828, | |
| "loss": 3.7461, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 2.4156963204471356, | |
| "grad_norm": 0.3439953327178955, | |
| "learning_rate": 0.0005713488642981945, | |
| "loss": 3.7458, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 2.4302515137401026, | |
| "grad_norm": 0.3205753266811371, | |
| "learning_rate": 0.0005711741409435061, | |
| "loss": 3.7441, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 2.4448067070330692, | |
| "grad_norm": 0.3132123053073883, | |
| "learning_rate": 0.0005709994175888177, | |
| "loss": 3.7499, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.4593619003260363, | |
| "grad_norm": 0.32091590762138367, | |
| "learning_rate": 0.0005708246942341292, | |
| "loss": 3.7432, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 2.4739170936190034, | |
| "grad_norm": 0.304622083902359, | |
| "learning_rate": 0.0005706499708794409, | |
| "loss": 3.7417, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.4884722869119704, | |
| "grad_norm": 0.3125957250595093, | |
| "learning_rate": 0.0005704752475247524, | |
| "loss": 3.7431, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 2.503027480204937, | |
| "grad_norm": 0.3097993731498718, | |
| "learning_rate": 0.0005703005241700641, | |
| "loss": 3.7525, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.517582673497904, | |
| "grad_norm": 0.30082228779792786, | |
| "learning_rate": 0.0005701258008153756, | |
| "loss": 3.7402, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 2.5321378667908707, | |
| "grad_norm": 0.3065928816795349, | |
| "learning_rate": 0.0005699510774606872, | |
| "loss": 3.7398, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 2.546693060083838, | |
| "grad_norm": 0.3146657645702362, | |
| "learning_rate": 0.0005697763541059988, | |
| "loss": 3.7402, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 2.561248253376805, | |
| "grad_norm": 0.3063797652721405, | |
| "learning_rate": 0.0005696016307513103, | |
| "loss": 3.7394, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.575803446669772, | |
| "grad_norm": 0.3362005054950714, | |
| "learning_rate": 0.000569426907396622, | |
| "loss": 3.747, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 2.590358639962739, | |
| "grad_norm": 0.32363709807395935, | |
| "learning_rate": 0.0005692521840419336, | |
| "loss": 3.734, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 2.6049138332557056, | |
| "grad_norm": 0.3284408450126648, | |
| "learning_rate": 0.0005690774606872452, | |
| "loss": 3.7302, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "grad_norm": 0.31467774510383606, | |
| "learning_rate": 0.0005689027373325567, | |
| "loss": 3.7282, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "eval_accuracy": 0.34779338370631996, | |
| "eval_loss": 3.7500975131988525, | |
| "eval_runtime": 83.7301, | |
| "eval_samples_per_second": 198.865, | |
| "eval_steps_per_second": 12.433, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.6340242198416393, | |
| "grad_norm": 0.33911052346229553, | |
| "learning_rate": 0.0005687280139778683, | |
| "loss": 3.7302, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 2.6485794131346063, | |
| "grad_norm": 0.2991912364959717, | |
| "learning_rate": 0.0005685532906231799, | |
| "loss": 3.7275, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 2.6631346064275734, | |
| "grad_norm": 0.3125496506690979, | |
| "learning_rate": 0.0005683785672684915, | |
| "loss": 3.722, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 2.6776897997205404, | |
| "grad_norm": 0.31298065185546875, | |
| "learning_rate": 0.0005682038439138031, | |
| "loss": 3.7288, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.692244993013507, | |
| "grad_norm": 0.317345529794693, | |
| "learning_rate": 0.0005680291205591147, | |
| "loss": 3.719, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 2.706800186306474, | |
| "grad_norm": 0.3251684010028839, | |
| "learning_rate": 0.0005678543972044263, | |
| "loss": 3.7243, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 2.721355379599441, | |
| "grad_norm": 0.30273106694221497, | |
| "learning_rate": 0.0005676796738497378, | |
| "loss": 3.7264, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 2.735910572892408, | |
| "grad_norm": 0.32294148206710815, | |
| "learning_rate": 0.0005675049504950495, | |
| "loss": 3.733, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.750465766185375, | |
| "grad_norm": 0.3247222900390625, | |
| "learning_rate": 0.0005673302271403611, | |
| "loss": 3.7364, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 2.765020959478342, | |
| "grad_norm": 0.29552945494651794, | |
| "learning_rate": 0.0005671555037856726, | |
| "loss": 3.7295, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.779576152771309, | |
| "grad_norm": 0.32103094458580017, | |
| "learning_rate": 0.0005669807804309842, | |
| "loss": 3.7169, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 2.7941313460642756, | |
| "grad_norm": 0.31306442618370056, | |
| "learning_rate": 0.0005668060570762958, | |
| "loss": 3.7269, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 2.8086865393572427, | |
| "grad_norm": 0.3265579044818878, | |
| "learning_rate": 0.0005666313337216074, | |
| "loss": 3.7257, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 2.8232417326502097, | |
| "grad_norm": 0.34516066312789917, | |
| "learning_rate": 0.000566456610366919, | |
| "loss": 3.7212, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 2.8377969259431763, | |
| "grad_norm": 0.31213825941085815, | |
| "learning_rate": 0.0005662818870122306, | |
| "loss": 3.7321, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 2.8523521192361434, | |
| "grad_norm": 0.29077741503715515, | |
| "learning_rate": 0.0005661071636575422, | |
| "loss": 3.7243, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 2.8669073125291105, | |
| "grad_norm": 0.30966341495513916, | |
| "learning_rate": 0.0005659324403028537, | |
| "loss": 3.7277, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 2.8814625058220775, | |
| "grad_norm": 0.31948623061180115, | |
| "learning_rate": 0.0005657577169481653, | |
| "loss": 3.7114, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 2.896017699115044, | |
| "grad_norm": 0.2940485179424286, | |
| "learning_rate": 0.0005655829935934769, | |
| "loss": 3.7012, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "grad_norm": 0.30057135224342346, | |
| "learning_rate": 0.0005654082702387886, | |
| "loss": 3.7136, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.910572892408011, | |
| "eval_accuracy": 0.3500327489836532, | |
| "eval_loss": 3.7230215072631836, | |
| "eval_runtime": 83.8309, | |
| "eval_samples_per_second": 198.626, | |
| "eval_steps_per_second": 12.418, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.9251280857009783, | |
| "grad_norm": 0.3131848871707916, | |
| "learning_rate": 0.0005652335468841001, | |
| "loss": 3.725, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 2.939683278993945, | |
| "grad_norm": 0.30893978476524353, | |
| "learning_rate": 0.0005650588235294117, | |
| "loss": 3.7017, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 2.954238472286912, | |
| "grad_norm": 0.32823804020881653, | |
| "learning_rate": 0.0005648841001747233, | |
| "loss": 3.7219, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 2.968793665579879, | |
| "grad_norm": 0.3186146914958954, | |
| "learning_rate": 0.0005647093768200349, | |
| "loss": 3.7241, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 2.983348858872846, | |
| "grad_norm": 0.3129662275314331, | |
| "learning_rate": 0.0005645346534653465, | |
| "loss": 3.7231, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 2.9979040521658127, | |
| "grad_norm": 0.3196828067302704, | |
| "learning_rate": 0.0005643599301106582, | |
| "loss": 3.7139, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 3.0122263623660923, | |
| "grad_norm": 0.33347222208976746, | |
| "learning_rate": 0.0005641852067559697, | |
| "loss": 3.6365, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 3.026781555659059, | |
| "grad_norm": 0.31537434458732605, | |
| "learning_rate": 0.0005640104834012812, | |
| "loss": 3.6007, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 3.041336748952026, | |
| "grad_norm": 0.33573049306869507, | |
| "learning_rate": 0.0005638357600465929, | |
| "loss": 3.614, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 3.055891942244993, | |
| "grad_norm": 0.32463812828063965, | |
| "learning_rate": 0.0005636610366919044, | |
| "loss": 3.6261, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.07044713553796, | |
| "grad_norm": 0.3028387427330017, | |
| "learning_rate": 0.0005634863133372161, | |
| "loss": 3.6131, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 3.0850023288309267, | |
| "grad_norm": 0.2982977628707886, | |
| "learning_rate": 0.0005633115899825276, | |
| "loss": 3.6232, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 3.099557522123894, | |
| "grad_norm": 0.3213280737400055, | |
| "learning_rate": 0.0005631368666278393, | |
| "loss": 3.6213, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 3.114112715416861, | |
| "grad_norm": 0.33530834317207336, | |
| "learning_rate": 0.0005629621432731508, | |
| "loss": 3.6191, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 3.1286679087098275, | |
| "grad_norm": 0.3106696903705597, | |
| "learning_rate": 0.0005627874199184623, | |
| "loss": 3.6332, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 3.1432231020027945, | |
| "grad_norm": 0.32566139101982117, | |
| "learning_rate": 0.000562612696563774, | |
| "loss": 3.6273, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 3.1577782952957616, | |
| "grad_norm": 0.324490487575531, | |
| "learning_rate": 0.0005624379732090856, | |
| "loss": 3.63, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 3.1723334885887287, | |
| "grad_norm": 0.3148472309112549, | |
| "learning_rate": 0.0005622632498543972, | |
| "loss": 3.6343, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 3.1868886818816953, | |
| "grad_norm": 0.3184763789176941, | |
| "learning_rate": 0.0005620885264997087, | |
| "loss": 3.6294, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 3.2014438751746623, | |
| "grad_norm": 0.33171960711479187, | |
| "learning_rate": 0.0005619138031450204, | |
| "loss": 3.627, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2014438751746623, | |
| "eval_accuracy": 0.35198041148895226, | |
| "eval_loss": 3.709048271179199, | |
| "eval_runtime": 83.8118, | |
| "eval_samples_per_second": 198.671, | |
| "eval_steps_per_second": 12.421, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.2159990684676294, | |
| "grad_norm": 0.3211822807788849, | |
| "learning_rate": 0.0005617390797903319, | |
| "loss": 3.6298, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 3.230554261760596, | |
| "grad_norm": 0.3232152760028839, | |
| "learning_rate": 0.0005615643564356436, | |
| "loss": 3.6407, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 3.245109455053563, | |
| "grad_norm": 0.32056182622909546, | |
| "learning_rate": 0.0005613896330809551, | |
| "loss": 3.6313, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 3.25966464834653, | |
| "grad_norm": 0.32539117336273193, | |
| "learning_rate": 0.0005612149097262667, | |
| "loss": 3.6422, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 3.274219841639497, | |
| "grad_norm": 0.3430807590484619, | |
| "learning_rate": 0.0005610401863715783, | |
| "loss": 3.6389, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 3.288775034932464, | |
| "grad_norm": 0.3201347887516022, | |
| "learning_rate": 0.0005608654630168898, | |
| "loss": 3.6181, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 3.303330228225431, | |
| "grad_norm": 0.33122941851615906, | |
| "learning_rate": 0.0005606907396622015, | |
| "loss": 3.6372, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 3.317885421518398, | |
| "grad_norm": 0.2859053313732147, | |
| "learning_rate": 0.0005605160163075131, | |
| "loss": 3.6334, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 3.3324406148113646, | |
| "grad_norm": 0.3348695933818817, | |
| "learning_rate": 0.0005603412929528247, | |
| "loss": 3.6289, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 3.3469958081043316, | |
| "grad_norm": 0.32294225692749023, | |
| "learning_rate": 0.0005601665695981362, | |
| "loss": 3.6381, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.3615510013972987, | |
| "grad_norm": 0.3325670063495636, | |
| "learning_rate": 0.0005599918462434478, | |
| "loss": 3.6287, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 3.3761061946902653, | |
| "grad_norm": 0.30032414197921753, | |
| "learning_rate": 0.0005598171228887594, | |
| "loss": 3.6282, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 3.3906613879832324, | |
| "grad_norm": 0.3063546121120453, | |
| "learning_rate": 0.0005596423995340709, | |
| "loss": 3.6411, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 3.4052165812761994, | |
| "grad_norm": 0.31318849325180054, | |
| "learning_rate": 0.0005594676761793826, | |
| "loss": 3.6545, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 3.419771774569166, | |
| "grad_norm": 0.3217429518699646, | |
| "learning_rate": 0.0005592929528246942, | |
| "loss": 3.6273, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 3.434326967862133, | |
| "grad_norm": 0.3252391517162323, | |
| "learning_rate": 0.0005591182294700058, | |
| "loss": 3.6299, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 3.4488821611551, | |
| "grad_norm": 0.31127187609672546, | |
| "learning_rate": 0.0005589435061153173, | |
| "loss": 3.6371, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 3.463437354448067, | |
| "grad_norm": 0.3100443184375763, | |
| "learning_rate": 0.000558768782760629, | |
| "loss": 3.6215, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 3.477992547741034, | |
| "grad_norm": 0.30956730246543884, | |
| "learning_rate": 0.0005585940594059406, | |
| "loss": 3.6446, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 3.492547741034001, | |
| "grad_norm": 0.29372090101242065, | |
| "learning_rate": 0.0005584193360512521, | |
| "loss": 3.6246, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.492547741034001, | |
| "eval_accuracy": 0.3539401793067088, | |
| "eval_loss": 3.6921117305755615, | |
| "eval_runtime": 83.8029, | |
| "eval_samples_per_second": 198.693, | |
| "eval_steps_per_second": 12.422, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.507102934326968, | |
| "grad_norm": 0.31082335114479065, | |
| "learning_rate": 0.0005582446126965637, | |
| "loss": 3.6426, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 3.5216581276199346, | |
| "grad_norm": 0.3189781904220581, | |
| "learning_rate": 0.0005580698893418753, | |
| "loss": 3.626, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 3.5362133209129016, | |
| "grad_norm": 0.3097034692764282, | |
| "learning_rate": 0.0005578951659871869, | |
| "loss": 3.6325, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 3.5507685142058687, | |
| "grad_norm": 0.3243505656719208, | |
| "learning_rate": 0.0005577204426324985, | |
| "loss": 3.6456, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 3.5653237074988358, | |
| "grad_norm": 0.31229543685913086, | |
| "learning_rate": 0.0005575457192778101, | |
| "loss": 3.6334, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 3.5798789007918024, | |
| "grad_norm": 0.3169929087162018, | |
| "learning_rate": 0.0005573709959231217, | |
| "loss": 3.6522, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 3.5944340940847694, | |
| "grad_norm": 0.3102095425128937, | |
| "learning_rate": 0.0005571962725684332, | |
| "loss": 3.6421, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 3.6089892873777365, | |
| "grad_norm": 0.3058861494064331, | |
| "learning_rate": 0.0005570215492137449, | |
| "loss": 3.6425, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 3.623544480670703, | |
| "grad_norm": 0.31340351700782776, | |
| "learning_rate": 0.0005568468258590564, | |
| "loss": 3.6248, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 3.63809967396367, | |
| "grad_norm": 0.3283979594707489, | |
| "learning_rate": 0.0005566721025043681, | |
| "loss": 3.6307, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.6526548672566372, | |
| "grad_norm": 0.30769020318984985, | |
| "learning_rate": 0.0005564973791496796, | |
| "loss": 3.627, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 3.6672100605496043, | |
| "grad_norm": 0.33009010553359985, | |
| "learning_rate": 0.0005563226557949913, | |
| "loss": 3.6451, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 3.681765253842571, | |
| "grad_norm": 0.31518447399139404, | |
| "learning_rate": 0.0005561479324403028, | |
| "loss": 3.6412, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 3.696320447135538, | |
| "grad_norm": 0.32417356967926025, | |
| "learning_rate": 0.0005559732090856144, | |
| "loss": 3.6428, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 3.710875640428505, | |
| "grad_norm": 0.307077020406723, | |
| "learning_rate": 0.000555798485730926, | |
| "loss": 3.6331, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 3.7254308337214717, | |
| "grad_norm": 0.31680402159690857, | |
| "learning_rate": 0.0005556237623762376, | |
| "loss": 3.635, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 3.7399860270144387, | |
| "grad_norm": 0.31820669770240784, | |
| "learning_rate": 0.0005554490390215492, | |
| "loss": 3.638, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 3.754541220307406, | |
| "grad_norm": 0.3132565915584564, | |
| "learning_rate": 0.0005552743156668607, | |
| "loss": 3.6474, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 3.769096413600373, | |
| "grad_norm": 0.3104014992713928, | |
| "learning_rate": 0.0005550995923121724, | |
| "loss": 3.6306, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 3.7836516068933395, | |
| "grad_norm": 0.3177199363708496, | |
| "learning_rate": 0.0005549248689574839, | |
| "loss": 3.6345, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.7836516068933395, | |
| "eval_accuracy": 0.3555451321894244, | |
| "eval_loss": 3.6750457286834717, | |
| "eval_runtime": 83.7353, | |
| "eval_samples_per_second": 198.853, | |
| "eval_steps_per_second": 12.432, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.7982068001863065, | |
| "grad_norm": 0.30314624309539795, | |
| "learning_rate": 0.0005547501456027955, | |
| "loss": 3.6447, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 3.812761993479273, | |
| "grad_norm": 0.2967151403427124, | |
| "learning_rate": 0.0005545754222481071, | |
| "loss": 3.6423, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 3.82731718677224, | |
| "grad_norm": 0.30352118611335754, | |
| "learning_rate": 0.0005544006988934188, | |
| "loss": 3.6301, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 3.8418723800652073, | |
| "grad_norm": 0.30857136845588684, | |
| "learning_rate": 0.0005542259755387303, | |
| "loss": 3.6353, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 3.8564275733581743, | |
| "grad_norm": 0.3036711812019348, | |
| "learning_rate": 0.0005540512521840418, | |
| "loss": 3.6244, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 3.8709827666511414, | |
| "grad_norm": 0.3091478943824768, | |
| "learning_rate": 0.0005538765288293535, | |
| "loss": 3.6255, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 3.885537959944108, | |
| "grad_norm": 0.3349607288837433, | |
| "learning_rate": 0.0005537018054746651, | |
| "loss": 3.635, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 3.900093153237075, | |
| "grad_norm": 0.3055279552936554, | |
| "learning_rate": 0.0005535270821199767, | |
| "loss": 3.6372, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 3.9146483465300417, | |
| "grad_norm": 0.31592175364494324, | |
| "learning_rate": 0.0005533523587652882, | |
| "loss": 3.6373, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 3.9292035398230087, | |
| "grad_norm": 0.3077903687953949, | |
| "learning_rate": 0.0005531776354105999, | |
| "loss": 3.6222, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.943758733115976, | |
| "grad_norm": 0.31184664368629456, | |
| "learning_rate": 0.0005530029120559114, | |
| "loss": 3.6392, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 3.958313926408943, | |
| "grad_norm": 0.3033412992954254, | |
| "learning_rate": 0.0005528281887012229, | |
| "loss": 3.6263, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 3.9728691197019095, | |
| "grad_norm": 0.3176705241203308, | |
| "learning_rate": 0.0005526534653465346, | |
| "loss": 3.6384, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 3.9874243129948765, | |
| "grad_norm": 0.2997167110443115, | |
| "learning_rate": 0.0005524787419918462, | |
| "loss": 3.6333, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 4.001746623195156, | |
| "grad_norm": 0.34342196583747864, | |
| "learning_rate": 0.0005523040186371578, | |
| "loss": 3.6161, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 4.016301816488123, | |
| "grad_norm": 0.31291690468788147, | |
| "learning_rate": 0.0005521292952824693, | |
| "loss": 3.5231, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 4.03085700978109, | |
| "grad_norm": 0.30333396792411804, | |
| "learning_rate": 0.000551954571927781, | |
| "loss": 3.5348, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 4.045412203074057, | |
| "grad_norm": 0.3121480941772461, | |
| "learning_rate": 0.0005517798485730926, | |
| "loss": 3.5243, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 4.059967396367024, | |
| "grad_norm": 0.32029443979263306, | |
| "learning_rate": 0.0005516051252184042, | |
| "loss": 3.5248, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 4.074522589659991, | |
| "grad_norm": 0.2967100739479065, | |
| "learning_rate": 0.0005514304018637157, | |
| "loss": 3.5324, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.074522589659991, | |
| "eval_accuracy": 0.3568756588140014, | |
| "eval_loss": 3.6658079624176025, | |
| "eval_runtime": 83.8295, | |
| "eval_samples_per_second": 198.629, | |
| "eval_steps_per_second": 12.418, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.089077782952957, | |
| "grad_norm": 0.31021130084991455, | |
| "learning_rate": 0.0005512556785090273, | |
| "loss": 3.5482, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 4.103632976245924, | |
| "grad_norm": 0.3215734362602234, | |
| "learning_rate": 0.0005510809551543389, | |
| "loss": 3.5397, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 4.118188169538891, | |
| "grad_norm": 0.318257212638855, | |
| "learning_rate": 0.0005509062317996504, | |
| "loss": 3.5383, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 4.132743362831858, | |
| "grad_norm": 0.31577590107917786, | |
| "learning_rate": 0.0005507315084449621, | |
| "loss": 3.5527, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 4.1472985561248255, | |
| "grad_norm": 0.32432398200035095, | |
| "learning_rate": 0.0005505567850902737, | |
| "loss": 3.5478, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 4.1618537494177925, | |
| "grad_norm": 0.3222511112689972, | |
| "learning_rate": 0.0005503820617355853, | |
| "loss": 3.5532, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 4.17640894271076, | |
| "grad_norm": 0.3277701139450073, | |
| "learning_rate": 0.0005502073383808969, | |
| "loss": 3.5555, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 4.190964136003726, | |
| "grad_norm": 0.3025704026222229, | |
| "learning_rate": 0.0005500326150262085, | |
| "loss": 3.5675, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 4.205519329296693, | |
| "grad_norm": 0.32261499762535095, | |
| "learning_rate": 0.00054985789167152, | |
| "loss": 3.5507, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 4.22007452258966, | |
| "grad_norm": 0.3120281398296356, | |
| "learning_rate": 0.0005496831683168316, | |
| "loss": 3.5602, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.234629715882627, | |
| "grad_norm": 0.3225678503513336, | |
| "learning_rate": 0.0005495084449621433, | |
| "loss": 3.5494, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 4.249184909175594, | |
| "grad_norm": 0.3101062774658203, | |
| "learning_rate": 0.0005493337216074548, | |
| "loss": 3.5554, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 4.263740102468561, | |
| "grad_norm": 0.3139819800853729, | |
| "learning_rate": 0.0005491589982527664, | |
| "loss": 3.5555, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 4.278295295761528, | |
| "grad_norm": 0.33591219782829285, | |
| "learning_rate": 0.000548984274898078, | |
| "loss": 3.5452, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 4.292850489054494, | |
| "grad_norm": 0.3268113434314728, | |
| "learning_rate": 0.0005488095515433897, | |
| "loss": 3.5519, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 4.307405682347461, | |
| "grad_norm": 0.3381194472312927, | |
| "learning_rate": 0.0005486348281887012, | |
| "loss": 3.5493, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 4.321960875640428, | |
| "grad_norm": 0.3033466041088104, | |
| "learning_rate": 0.0005484601048340127, | |
| "loss": 3.5631, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 4.3365160689333955, | |
| "grad_norm": 0.32271185517311096, | |
| "learning_rate": 0.0005482853814793244, | |
| "loss": 3.5574, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 4.3510712622263625, | |
| "grad_norm": 0.31459832191467285, | |
| "learning_rate": 0.0005481106581246359, | |
| "loss": 3.5654, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 4.36562645551933, | |
| "grad_norm": 0.32392099499702454, | |
| "learning_rate": 0.0005479359347699475, | |
| "loss": 3.5519, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.36562645551933, | |
| "eval_accuracy": 0.35767825278266463, | |
| "eval_loss": 3.6543850898742676, | |
| "eval_runtime": 83.7224, | |
| "eval_samples_per_second": 198.884, | |
| "eval_steps_per_second": 12.434, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.380181648812297, | |
| "grad_norm": 0.32713738083839417, | |
| "learning_rate": 0.0005477612114152591, | |
| "loss": 3.5577, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 4.394736842105263, | |
| "grad_norm": 0.3144247233867645, | |
| "learning_rate": 0.0005475864880605708, | |
| "loss": 3.56, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 4.40929203539823, | |
| "grad_norm": 0.31902626156806946, | |
| "learning_rate": 0.0005474117647058823, | |
| "loss": 3.5707, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 4.423847228691197, | |
| "grad_norm": 0.28951430320739746, | |
| "learning_rate": 0.0005472370413511939, | |
| "loss": 3.5578, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 4.438402421984164, | |
| "grad_norm": 0.322680801153183, | |
| "learning_rate": 0.0005470623179965055, | |
| "loss": 3.5766, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 4.452957615277131, | |
| "grad_norm": 0.3243967294692993, | |
| "learning_rate": 0.0005468875946418171, | |
| "loss": 3.5731, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 4.467512808570098, | |
| "grad_norm": 0.314889132976532, | |
| "learning_rate": 0.0005467128712871287, | |
| "loss": 3.5704, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 4.482068001863064, | |
| "grad_norm": 0.3427557647228241, | |
| "learning_rate": 0.0005465381479324402, | |
| "loss": 3.5564, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 4.496623195156031, | |
| "grad_norm": 0.3235917389392853, | |
| "learning_rate": 0.0005463634245777519, | |
| "loss": 3.5645, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 4.511178388448998, | |
| "grad_norm": 0.3178023397922516, | |
| "learning_rate": 0.0005461887012230634, | |
| "loss": 3.5567, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 4.5257335817419655, | |
| "grad_norm": 0.30542704463005066, | |
| "learning_rate": 0.000546013977868375, | |
| "loss": 3.5627, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 4.5402887750349326, | |
| "grad_norm": 0.29802826046943665, | |
| "learning_rate": 0.0005458392545136866, | |
| "loss": 3.5901, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 4.5548439683279, | |
| "grad_norm": 0.338664174079895, | |
| "learning_rate": 0.0005456645311589983, | |
| "loss": 3.5714, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 4.569399161620867, | |
| "grad_norm": 0.3050200045108795, | |
| "learning_rate": 0.0005454898078043098, | |
| "loss": 3.561, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 4.583954354913834, | |
| "grad_norm": 0.31070205569267273, | |
| "learning_rate": 0.0005453150844496213, | |
| "loss": 3.5667, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 4.5985095482068, | |
| "grad_norm": 0.29910558462142944, | |
| "learning_rate": 0.000545140361094933, | |
| "loss": 3.5746, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 4.613064741499767, | |
| "grad_norm": 0.3532266616821289, | |
| "learning_rate": 0.0005449656377402445, | |
| "loss": 3.5665, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 4.627619934792734, | |
| "grad_norm": 0.33198997378349304, | |
| "learning_rate": 0.0005447909143855562, | |
| "loss": 3.5812, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 4.642175128085701, | |
| "grad_norm": 0.3118351995944977, | |
| "learning_rate": 0.0005446161910308677, | |
| "loss": 3.5728, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 4.656730321378668, | |
| "grad_norm": 0.30612584948539734, | |
| "learning_rate": 0.0005444414676761794, | |
| "loss": 3.5647, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.656730321378668, | |
| "eval_accuracy": 0.3592765066089717, | |
| "eval_loss": 3.640305519104004, | |
| "eval_runtime": 83.8647, | |
| "eval_samples_per_second": 198.546, | |
| "eval_steps_per_second": 12.413, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.671285514671635, | |
| "grad_norm": 0.3129040598869324, | |
| "learning_rate": 0.0005442667443214909, | |
| "loss": 3.5666, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 4.685840707964601, | |
| "grad_norm": 0.3124503195285797, | |
| "learning_rate": 0.0005440920209668024, | |
| "loss": 3.5643, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 4.7003959012575685, | |
| "grad_norm": 0.3141111433506012, | |
| "learning_rate": 0.0005439172976121141, | |
| "loss": 3.5771, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 4.7149510945505355, | |
| "grad_norm": 0.3096890151500702, | |
| "learning_rate": 0.0005437425742574257, | |
| "loss": 3.5758, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 4.729506287843503, | |
| "grad_norm": 0.328483521938324, | |
| "learning_rate": 0.0005435678509027373, | |
| "loss": 3.5615, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 4.74406148113647, | |
| "grad_norm": 0.3097730278968811, | |
| "learning_rate": 0.0005433931275480488, | |
| "loss": 3.5607, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 4.758616674429437, | |
| "grad_norm": 0.29414355754852295, | |
| "learning_rate": 0.0005432184041933605, | |
| "loss": 3.5633, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 4.773171867722404, | |
| "grad_norm": 0.3001692295074463, | |
| "learning_rate": 0.000543043680838672, | |
| "loss": 3.5618, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 4.78772706101537, | |
| "grad_norm": 0.29483455419540405, | |
| "learning_rate": 0.0005428689574839837, | |
| "loss": 3.5766, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 4.802282254308337, | |
| "grad_norm": 0.3080602288246155, | |
| "learning_rate": 0.0005426942341292952, | |
| "loss": 3.5704, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 4.816837447601304, | |
| "grad_norm": 0.32206419110298157, | |
| "learning_rate": 0.0005425195107746068, | |
| "loss": 3.5654, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 4.831392640894271, | |
| "grad_norm": 0.3105108141899109, | |
| "learning_rate": 0.0005423447874199184, | |
| "loss": 3.5659, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 4.845947834187238, | |
| "grad_norm": 0.3111788332462311, | |
| "learning_rate": 0.00054217006406523, | |
| "loss": 3.571, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 4.860503027480205, | |
| "grad_norm": 0.3317568600177765, | |
| "learning_rate": 0.0005419953407105417, | |
| "loss": 3.5732, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 4.875058220773171, | |
| "grad_norm": 0.30133309960365295, | |
| "learning_rate": 0.0005418206173558532, | |
| "loss": 3.5685, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 4.8896134140661385, | |
| "grad_norm": 0.31733211874961853, | |
| "learning_rate": 0.0005416458940011648, | |
| "loss": 3.5665, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 4.9041686073591055, | |
| "grad_norm": 0.31207114458084106, | |
| "learning_rate": 0.0005414711706464764, | |
| "loss": 3.5706, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 4.918723800652073, | |
| "grad_norm": 0.31166326999664307, | |
| "learning_rate": 0.000541296447291788, | |
| "loss": 3.5748, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 4.93327899394504, | |
| "grad_norm": 0.29991692304611206, | |
| "learning_rate": 0.0005411217239370995, | |
| "loss": 3.5719, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 4.947834187238007, | |
| "grad_norm": 0.3093617558479309, | |
| "learning_rate": 0.0005409470005824111, | |
| "loss": 3.5804, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.947834187238007, | |
| "eval_accuracy": 0.36033295955732636, | |
| "eval_loss": 3.6289567947387695, | |
| "eval_runtime": 83.7757, | |
| "eval_samples_per_second": 198.757, | |
| "eval_steps_per_second": 12.426, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.962389380530974, | |
| "grad_norm": 0.29420968890190125, | |
| "learning_rate": 0.0005407722772277228, | |
| "loss": 3.5827, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 4.976944573823941, | |
| "grad_norm": 0.3155781924724579, | |
| "learning_rate": 0.0005405975538730343, | |
| "loss": 3.577, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 4.991499767116907, | |
| "grad_norm": 0.31461986899375916, | |
| "learning_rate": 0.0005404228305183459, | |
| "loss": 3.5703, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 5.005822077317187, | |
| "grad_norm": 0.32665446400642395, | |
| "learning_rate": 0.0005402481071636575, | |
| "loss": 3.5333, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 5.020377270610154, | |
| "grad_norm": 0.3217109143733978, | |
| "learning_rate": 0.0005400733838089692, | |
| "loss": 3.4542, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 5.034932463903121, | |
| "grad_norm": 0.31351569294929504, | |
| "learning_rate": 0.0005398986604542807, | |
| "loss": 3.4518, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 5.049487657196088, | |
| "grad_norm": 0.3086356818675995, | |
| "learning_rate": 0.0005397239370995922, | |
| "loss": 3.4642, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 5.064042850489055, | |
| "grad_norm": 0.31107228994369507, | |
| "learning_rate": 0.0005395492137449039, | |
| "loss": 3.4773, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 5.078598043782021, | |
| "grad_norm": 0.3050672709941864, | |
| "learning_rate": 0.0005393744903902154, | |
| "loss": 3.4795, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 5.093153237074988, | |
| "grad_norm": 0.3185788691043854, | |
| "learning_rate": 0.000539199767035527, | |
| "loss": 3.4747, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 5.107708430367955, | |
| "grad_norm": 0.33891022205352783, | |
| "learning_rate": 0.0005390250436808386, | |
| "loss": 3.4781, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 5.122263623660922, | |
| "grad_norm": 0.3091220259666443, | |
| "learning_rate": 0.0005388503203261503, | |
| "loss": 3.4796, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 5.136818816953889, | |
| "grad_norm": 0.31760117411613464, | |
| "learning_rate": 0.0005386755969714618, | |
| "loss": 3.4901, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 5.151374010246856, | |
| "grad_norm": 0.3264091908931732, | |
| "learning_rate": 0.0005385008736167733, | |
| "loss": 3.4796, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 5.165929203539823, | |
| "grad_norm": 0.31651178002357483, | |
| "learning_rate": 0.000538326150262085, | |
| "loss": 3.4911, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 5.18048439683279, | |
| "grad_norm": 0.33168500661849976, | |
| "learning_rate": 0.0005381514269073965, | |
| "loss": 3.4957, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 5.195039590125757, | |
| "grad_norm": 0.31706807017326355, | |
| "learning_rate": 0.0005379767035527082, | |
| "loss": 3.4871, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 5.209594783418724, | |
| "grad_norm": 0.31430307030677795, | |
| "learning_rate": 0.0005378019801980197, | |
| "loss": 3.4863, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 5.224149976711691, | |
| "grad_norm": 0.3324972093105316, | |
| "learning_rate": 0.0005376272568433314, | |
| "loss": 3.4952, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 5.238705170004658, | |
| "grad_norm": 0.30846595764160156, | |
| "learning_rate": 0.0005374525334886429, | |
| "loss": 3.495, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.238705170004658, | |
| "eval_accuracy": 0.36055156034539393, | |
| "eval_loss": 3.6321306228637695, | |
| "eval_runtime": 83.7964, | |
| "eval_samples_per_second": 198.708, | |
| "eval_steps_per_second": 12.423, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.253260363297625, | |
| "grad_norm": 0.3067502975463867, | |
| "learning_rate": 0.0005372778101339545, | |
| "loss": 3.4974, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 5.267815556590591, | |
| "grad_norm": 0.31356900930404663, | |
| "learning_rate": 0.0005371030867792661, | |
| "loss": 3.5117, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 5.282370749883558, | |
| "grad_norm": 0.3164156675338745, | |
| "learning_rate": 0.0005369283634245778, | |
| "loss": 3.4991, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 5.296925943176525, | |
| "grad_norm": 0.32259994745254517, | |
| "learning_rate": 0.0005367536400698893, | |
| "loss": 3.5036, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 5.311481136469492, | |
| "grad_norm": 0.32768815755844116, | |
| "learning_rate": 0.0005365789167152008, | |
| "loss": 3.5038, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 5.326036329762459, | |
| "grad_norm": 0.3055580258369446, | |
| "learning_rate": 0.0005364041933605125, | |
| "loss": 3.5054, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 5.340591523055426, | |
| "grad_norm": 0.3168846666812897, | |
| "learning_rate": 0.000536229470005824, | |
| "loss": 3.5152, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 5.3551467163483935, | |
| "grad_norm": 0.328299880027771, | |
| "learning_rate": 0.0005360547466511357, | |
| "loss": 3.5101, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 5.36970190964136, | |
| "grad_norm": 0.3162803649902344, | |
| "learning_rate": 0.0005358800232964472, | |
| "loss": 3.5149, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 5.384257102934327, | |
| "grad_norm": 0.3213299810886383, | |
| "learning_rate": 0.0005357052999417589, | |
| "loss": 3.5127, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 5.398812296227294, | |
| "grad_norm": 0.31621253490448, | |
| "learning_rate": 0.0005355305765870704, | |
| "loss": 3.5117, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 5.413367489520261, | |
| "grad_norm": 0.3191957473754883, | |
| "learning_rate": 0.000535355853232382, | |
| "loss": 3.517, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 5.427922682813228, | |
| "grad_norm": 0.3197634816169739, | |
| "learning_rate": 0.0005351811298776936, | |
| "loss": 3.5116, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 5.442477876106195, | |
| "grad_norm": 0.3146786689758301, | |
| "learning_rate": 0.0005350064065230052, | |
| "loss": 3.5036, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 5.457033069399162, | |
| "grad_norm": 0.3258400857448578, | |
| "learning_rate": 0.0005348316831683168, | |
| "loss": 3.5229, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 5.471588262692128, | |
| "grad_norm": 0.33452728390693665, | |
| "learning_rate": 0.0005346569598136284, | |
| "loss": 3.513, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 5.486143455985095, | |
| "grad_norm": 0.3063402771949768, | |
| "learning_rate": 0.00053448223645894, | |
| "loss": 3.5159, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 5.500698649278062, | |
| "grad_norm": 0.3065362572669983, | |
| "learning_rate": 0.0005343075131042515, | |
| "loss": 3.5095, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 5.515253842571029, | |
| "grad_norm": 0.316665381193161, | |
| "learning_rate": 0.0005341327897495632, | |
| "loss": 3.5093, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 5.529809035863996, | |
| "grad_norm": 0.3052302896976471, | |
| "learning_rate": 0.0005339580663948748, | |
| "loss": 3.5176, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.529809035863996, | |
| "eval_accuracy": 0.36188890355368486, | |
| "eval_loss": 3.6205837726593018, | |
| "eval_runtime": 83.729, | |
| "eval_samples_per_second": 198.868, | |
| "eval_steps_per_second": 12.433, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.5443642291569635, | |
| "grad_norm": 0.6032854914665222, | |
| "learning_rate": 0.0005337833430401863, | |
| "loss": 3.5087, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 5.5589194224499305, | |
| "grad_norm": 0.32861748337745667, | |
| "learning_rate": 0.0005336086196854979, | |
| "loss": 3.5218, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 5.573474615742897, | |
| "grad_norm": 0.32680070400238037, | |
| "learning_rate": 0.0005334338963308095, | |
| "loss": 3.5229, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 5.588029809035864, | |
| "grad_norm": 0.32045724987983704, | |
| "learning_rate": 0.0005332591729761211, | |
| "loss": 3.5273, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 5.602585002328831, | |
| "grad_norm": 0.310803085565567, | |
| "learning_rate": 0.0005330844496214327, | |
| "loss": 3.5205, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 5.617140195621798, | |
| "grad_norm": 0.3171592652797699, | |
| "learning_rate": 0.0005329097262667443, | |
| "loss": 3.5113, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 5.631695388914765, | |
| "grad_norm": 0.33249717950820923, | |
| "learning_rate": 0.0005327350029120559, | |
| "loss": 3.5223, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 5.646250582207732, | |
| "grad_norm": 0.31446245312690735, | |
| "learning_rate": 0.0005325602795573674, | |
| "loss": 3.5228, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 5.660805775500698, | |
| "grad_norm": 0.3157249391078949, | |
| "learning_rate": 0.000532385556202679, | |
| "loss": 3.5166, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 5.675360968793665, | |
| "grad_norm": 0.32851624488830566, | |
| "learning_rate": 0.0005322108328479906, | |
| "loss": 3.5261, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 5.689916162086632, | |
| "grad_norm": 0.30693915486335754, | |
| "learning_rate": 0.0005320361094933023, | |
| "loss": 3.5308, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 5.704471355379599, | |
| "grad_norm": 0.31030866503715515, | |
| "learning_rate": 0.0005318613861386138, | |
| "loss": 3.5202, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 5.719026548672566, | |
| "grad_norm": 0.31365323066711426, | |
| "learning_rate": 0.0005316866627839254, | |
| "loss": 3.5146, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 5.7335817419655335, | |
| "grad_norm": 0.32892748713493347, | |
| "learning_rate": 0.000531511939429237, | |
| "loss": 3.5181, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 5.748136935258501, | |
| "grad_norm": 0.3211446702480316, | |
| "learning_rate": 0.0005313372160745486, | |
| "loss": 3.5176, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 5.762692128551468, | |
| "grad_norm": 0.3176540732383728, | |
| "learning_rate": 0.0005311624927198602, | |
| "loss": 3.4995, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 5.777247321844434, | |
| "grad_norm": 0.34687095880508423, | |
| "learning_rate": 0.0005309877693651717, | |
| "loss": 3.5269, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 5.791802515137401, | |
| "grad_norm": 0.29734715819358826, | |
| "learning_rate": 0.0005308130460104834, | |
| "loss": 3.522, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 5.806357708430368, | |
| "grad_norm": 0.3133270740509033, | |
| "learning_rate": 0.0005306383226557949, | |
| "loss": 3.5218, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 5.820912901723335, | |
| "grad_norm": 0.3237532079219818, | |
| "learning_rate": 0.0005304635993011065, | |
| "loss": 3.5285, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.820912901723335, | |
| "eval_accuracy": 0.3626745735903687, | |
| "eval_loss": 3.6092092990875244, | |
| "eval_runtime": 83.7604, | |
| "eval_samples_per_second": 198.793, | |
| "eval_steps_per_second": 12.428, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 5.835468095016302, | |
| "grad_norm": 0.3210567831993103, | |
| "learning_rate": 0.0005302888759464181, | |
| "loss": 3.5313, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 5.850023288309269, | |
| "grad_norm": 0.32515671849250793, | |
| "learning_rate": 0.0005301141525917298, | |
| "loss": 3.5246, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 5.864578481602235, | |
| "grad_norm": 0.2956371605396271, | |
| "learning_rate": 0.0005299394292370413, | |
| "loss": 3.52, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 5.879133674895202, | |
| "grad_norm": 0.3496900200843811, | |
| "learning_rate": 0.0005297647058823528, | |
| "loss": 3.5179, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 5.893688868188169, | |
| "grad_norm": 0.3140091896057129, | |
| "learning_rate": 0.0005295899825276645, | |
| "loss": 3.5158, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 5.9082440614811365, | |
| "grad_norm": 0.310099720954895, | |
| "learning_rate": 0.000529415259172976, | |
| "loss": 3.5199, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 5.9227992547741035, | |
| "grad_norm": 0.32138338685035706, | |
| "learning_rate": 0.0005292405358182877, | |
| "loss": 3.5299, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 5.937354448067071, | |
| "grad_norm": 0.30885353684425354, | |
| "learning_rate": 0.0005290658124635992, | |
| "loss": 3.5248, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 5.951909641360038, | |
| "grad_norm": 0.29604068398475647, | |
| "learning_rate": 0.0005288910891089109, | |
| "loss": 3.5288, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 5.966464834653004, | |
| "grad_norm": 0.3159986138343811, | |
| "learning_rate": 0.0005287163657542224, | |
| "loss": 3.5242, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 5.981020027945971, | |
| "grad_norm": 0.31810715794563293, | |
| "learning_rate": 0.000528541642399534, | |
| "loss": 3.5292, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 5.995575221238938, | |
| "grad_norm": 0.3060857355594635, | |
| "learning_rate": 0.0005283669190448456, | |
| "loss": 3.5273, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 6.009897531439218, | |
| "grad_norm": 0.3063332438468933, | |
| "learning_rate": 0.0005281921956901572, | |
| "loss": 3.4461, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 6.024452724732185, | |
| "grad_norm": 0.3123130798339844, | |
| "learning_rate": 0.0005280174723354688, | |
| "loss": 3.4041, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 6.039007918025152, | |
| "grad_norm": 0.3306262493133545, | |
| "learning_rate": 0.0005278427489807804, | |
| "loss": 3.4264, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 6.053563111318118, | |
| "grad_norm": 0.3286329507827759, | |
| "learning_rate": 0.000527668025626092, | |
| "loss": 3.4094, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 6.068118304611085, | |
| "grad_norm": 0.3040468990802765, | |
| "learning_rate": 0.0005274933022714035, | |
| "loss": 3.433, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 6.082673497904052, | |
| "grad_norm": 0.31770357489585876, | |
| "learning_rate": 0.0005273185789167152, | |
| "loss": 3.4178, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 6.097228691197019, | |
| "grad_norm": 0.3256250321865082, | |
| "learning_rate": 0.0005271438555620268, | |
| "loss": 3.4441, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 6.111783884489986, | |
| "grad_norm": 0.33254656195640564, | |
| "learning_rate": 0.0005269691322073384, | |
| "loss": 3.4359, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.111783884489986, | |
| "eval_accuracy": 0.363104135891652, | |
| "eval_loss": 3.6126019954681396, | |
| "eval_runtime": 83.7358, | |
| "eval_samples_per_second": 198.852, | |
| "eval_steps_per_second": 12.432, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.126339077782953, | |
| "grad_norm": 0.3574708104133606, | |
| "learning_rate": 0.0005267944088526499, | |
| "loss": 3.4408, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 6.14089427107592, | |
| "grad_norm": 0.312255322933197, | |
| "learning_rate": 0.0005266196854979615, | |
| "loss": 3.4342, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 6.155449464368886, | |
| "grad_norm": 0.3316413164138794, | |
| "learning_rate": 0.0005264449621432731, | |
| "loss": 3.4516, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 6.1700046576618535, | |
| "grad_norm": 0.33468592166900635, | |
| "learning_rate": 0.0005262702387885847, | |
| "loss": 3.4449, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 6.1845598509548205, | |
| "grad_norm": 0.3310665190219879, | |
| "learning_rate": 0.0005260955154338963, | |
| "loss": 3.4462, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 6.199115044247788, | |
| "grad_norm": 0.31933286786079407, | |
| "learning_rate": 0.0005259207920792079, | |
| "loss": 3.4574, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 6.213670237540755, | |
| "grad_norm": 0.33896899223327637, | |
| "learning_rate": 0.0005257460687245195, | |
| "loss": 3.4572, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 6.228225430833722, | |
| "grad_norm": 0.3445572853088379, | |
| "learning_rate": 0.000525571345369831, | |
| "loss": 3.4622, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 6.242780624126689, | |
| "grad_norm": 0.2936427891254425, | |
| "learning_rate": 0.0005253966220151426, | |
| "loss": 3.4482, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 6.257335817419655, | |
| "grad_norm": 0.32076504826545715, | |
| "learning_rate": 0.0005252218986604543, | |
| "loss": 3.4619, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 6.271891010712622, | |
| "grad_norm": 0.3313353359699249, | |
| "learning_rate": 0.0005250471753057658, | |
| "loss": 3.45, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 6.286446204005589, | |
| "grad_norm": 0.3135939836502075, | |
| "learning_rate": 0.0005248724519510774, | |
| "loss": 3.4544, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 6.301001397298556, | |
| "grad_norm": 0.32245102524757385, | |
| "learning_rate": 0.000524697728596389, | |
| "loss": 3.4576, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 6.315556590591523, | |
| "grad_norm": 0.32054683566093445, | |
| "learning_rate": 0.0005245230052417006, | |
| "loss": 3.4468, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 6.33011178388449, | |
| "grad_norm": 0.3243516683578491, | |
| "learning_rate": 0.0005243482818870122, | |
| "loss": 3.4636, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 6.344666977177457, | |
| "grad_norm": 0.30357831716537476, | |
| "learning_rate": 0.0005241735585323238, | |
| "loss": 3.4589, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 6.3592221704704235, | |
| "grad_norm": 0.3101622760295868, | |
| "learning_rate": 0.0005239988351776354, | |
| "loss": 3.4566, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 6.3737773637633905, | |
| "grad_norm": 0.3167803883552551, | |
| "learning_rate": 0.0005238241118229469, | |
| "loss": 3.4738, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 6.388332557056358, | |
| "grad_norm": 0.3259917199611664, | |
| "learning_rate": 0.0005236493884682585, | |
| "loss": 3.4553, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 6.402887750349325, | |
| "grad_norm": 0.32265791296958923, | |
| "learning_rate": 0.0005234746651135701, | |
| "loss": 3.4691, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.402887750349325, | |
| "eval_accuracy": 0.3636048022127101, | |
| "eval_loss": 3.605232000350952, | |
| "eval_runtime": 83.8214, | |
| "eval_samples_per_second": 198.648, | |
| "eval_steps_per_second": 12.419, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.417442943642292, | |
| "grad_norm": 0.32583096623420715, | |
| "learning_rate": 0.0005232999417588818, | |
| "loss": 3.4637, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 6.431998136935259, | |
| "grad_norm": 0.3326973021030426, | |
| "learning_rate": 0.0005231252184041933, | |
| "loss": 3.4708, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 6.446553330228225, | |
| "grad_norm": 0.33622145652770996, | |
| "learning_rate": 0.0005229504950495049, | |
| "loss": 3.4558, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 6.461108523521192, | |
| "grad_norm": 0.3402341306209564, | |
| "learning_rate": 0.0005227757716948165, | |
| "loss": 3.4587, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 6.475663716814159, | |
| "grad_norm": 0.33382219076156616, | |
| "learning_rate": 0.000522601048340128, | |
| "loss": 3.4784, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 6.490218910107126, | |
| "grad_norm": 0.3161851167678833, | |
| "learning_rate": 0.0005224263249854397, | |
| "loss": 3.4691, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 6.504774103400093, | |
| "grad_norm": 0.34391483664512634, | |
| "learning_rate": 0.0005222516016307512, | |
| "loss": 3.4797, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 6.51932929669306, | |
| "grad_norm": 0.31955862045288086, | |
| "learning_rate": 0.0005220768782760629, | |
| "loss": 3.477, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 6.533884489986027, | |
| "grad_norm": 0.3096388876438141, | |
| "learning_rate": 0.0005219021549213744, | |
| "loss": 3.4712, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 6.548439683278994, | |
| "grad_norm": 0.3196328580379486, | |
| "learning_rate": 0.000521727431566686, | |
| "loss": 3.4777, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 6.562994876571961, | |
| "grad_norm": 0.3177974820137024, | |
| "learning_rate": 0.0005215527082119976, | |
| "loss": 3.4734, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 6.577550069864928, | |
| "grad_norm": 0.31900230050086975, | |
| "learning_rate": 0.0005213779848573093, | |
| "loss": 3.4777, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 6.592105263157895, | |
| "grad_norm": 0.32414501905441284, | |
| "learning_rate": 0.0005212032615026208, | |
| "loss": 3.4798, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 6.606660456450862, | |
| "grad_norm": 0.32272860407829285, | |
| "learning_rate": 0.0005210285381479323, | |
| "loss": 3.4741, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 6.621215649743829, | |
| "grad_norm": 0.35124295949935913, | |
| "learning_rate": 0.000520853814793244, | |
| "loss": 3.4786, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 6.635770843036796, | |
| "grad_norm": 0.3442808985710144, | |
| "learning_rate": 0.0005206790914385555, | |
| "loss": 3.4733, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 6.650326036329762, | |
| "grad_norm": 0.30354318022727966, | |
| "learning_rate": 0.0005205043680838672, | |
| "loss": 3.4776, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 6.664881229622729, | |
| "grad_norm": 0.3082331717014313, | |
| "learning_rate": 0.0005203296447291787, | |
| "loss": 3.4778, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 6.679436422915696, | |
| "grad_norm": 0.3416667878627777, | |
| "learning_rate": 0.0005201549213744904, | |
| "loss": 3.4892, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 6.693991616208663, | |
| "grad_norm": 0.3169821500778198, | |
| "learning_rate": 0.0005199801980198019, | |
| "loss": 3.4806, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.693991616208663, | |
| "eval_accuracy": 0.36465502621387785, | |
| "eval_loss": 3.595005989074707, | |
| "eval_runtime": 83.7962, | |
| "eval_samples_per_second": 198.708, | |
| "eval_steps_per_second": 12.423, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 6.70854680950163, | |
| "grad_norm": 0.3231274485588074, | |
| "learning_rate": 0.0005198054746651136, | |
| "loss": 3.4867, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 6.723102002794597, | |
| "grad_norm": 0.33539944887161255, | |
| "learning_rate": 0.0005196307513104251, | |
| "loss": 3.4778, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 6.737657196087564, | |
| "grad_norm": 0.32895907759666443, | |
| "learning_rate": 0.0005194560279557367, | |
| "loss": 3.4718, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 6.752212389380531, | |
| "grad_norm": 0.3291444182395935, | |
| "learning_rate": 0.0005192813046010483, | |
| "loss": 3.4866, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 6.766767582673498, | |
| "grad_norm": 0.30087754130363464, | |
| "learning_rate": 0.0005191065812463599, | |
| "loss": 3.4789, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 6.781322775966465, | |
| "grad_norm": 0.3096534311771393, | |
| "learning_rate": 0.0005189318578916715, | |
| "loss": 3.4836, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 6.795877969259432, | |
| "grad_norm": 0.3282308280467987, | |
| "learning_rate": 0.000518757134536983, | |
| "loss": 3.4929, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 6.810433162552399, | |
| "grad_norm": 0.3359972834587097, | |
| "learning_rate": 0.0005185824111822947, | |
| "loss": 3.4873, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 6.824988355845366, | |
| "grad_norm": 0.3327355682849884, | |
| "learning_rate": 0.0005184076878276063, | |
| "loss": 3.489, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 6.839543549138332, | |
| "grad_norm": 0.31763768196105957, | |
| "learning_rate": 0.0005182329644729179, | |
| "loss": 3.4807, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 6.854098742431299, | |
| "grad_norm": 0.31446197628974915, | |
| "learning_rate": 0.0005180582411182294, | |
| "loss": 3.4847, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 6.868653935724266, | |
| "grad_norm": 0.2972947359085083, | |
| "learning_rate": 0.000517883517763541, | |
| "loss": 3.4916, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 6.883209129017233, | |
| "grad_norm": 0.3304957449436188, | |
| "learning_rate": 0.0005177087944088526, | |
| "loss": 3.4923, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 6.8977643223102, | |
| "grad_norm": 0.3123498857021332, | |
| "learning_rate": 0.0005175340710541642, | |
| "loss": 3.4914, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 6.912319515603167, | |
| "grad_norm": 0.30580994486808777, | |
| "learning_rate": 0.0005173593476994758, | |
| "loss": 3.4835, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 6.926874708896134, | |
| "grad_norm": 0.30782079696655273, | |
| "learning_rate": 0.0005171846243447874, | |
| "loss": 3.5021, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 6.9414299021891015, | |
| "grad_norm": 0.3052419126033783, | |
| "learning_rate": 0.000517009900990099, | |
| "loss": 3.4831, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 6.955985095482068, | |
| "grad_norm": 0.32219356298446655, | |
| "learning_rate": 0.0005168351776354105, | |
| "loss": 3.4952, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 6.970540288775035, | |
| "grad_norm": 0.2996533513069153, | |
| "learning_rate": 0.0005166604542807221, | |
| "loss": 3.5055, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 6.985095482068002, | |
| "grad_norm": 0.3456665575504303, | |
| "learning_rate": 0.0005164857309260338, | |
| "loss": 3.4859, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.985095482068002, | |
| "eval_accuracy": 0.3649373268014791, | |
| "eval_loss": 3.5877349376678467, | |
| "eval_runtime": 83.7777, | |
| "eval_samples_per_second": 198.752, | |
| "eval_steps_per_second": 12.426, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 6.999650675360969, | |
| "grad_norm": 0.33315643668174744, | |
| "learning_rate": 0.0005163110075713453, | |
| "loss": 3.4849, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 7.0139729855612485, | |
| "grad_norm": 0.34144341945648193, | |
| "learning_rate": 0.0005161362842166569, | |
| "loss": 3.3809, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 7.0285281788542155, | |
| "grad_norm": 0.31783297657966614, | |
| "learning_rate": 0.0005159615608619685, | |
| "loss": 3.3776, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 7.043083372147182, | |
| "grad_norm": 0.3454691171646118, | |
| "learning_rate": 0.0005157868375072801, | |
| "loss": 3.3783, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 7.057638565440149, | |
| "grad_norm": 0.3108464181423187, | |
| "learning_rate": 0.0005156121141525917, | |
| "loss": 3.3897, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 7.072193758733116, | |
| "grad_norm": 0.33063140511512756, | |
| "learning_rate": 0.0005154373907979033, | |
| "loss": 3.3918, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 7.086748952026083, | |
| "grad_norm": 0.3560406267642975, | |
| "learning_rate": 0.0005152626674432149, | |
| "loss": 3.3965, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 7.10130414531905, | |
| "grad_norm": 0.3251829743385315, | |
| "learning_rate": 0.0005150879440885264, | |
| "loss": 3.3919, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 7.115859338612017, | |
| "grad_norm": 0.32970425486564636, | |
| "learning_rate": 0.000514913220733838, | |
| "loss": 3.3938, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 7.130414531904984, | |
| "grad_norm": 0.3270750939846039, | |
| "learning_rate": 0.0005147384973791496, | |
| "loss": 3.4123, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 7.14496972519795, | |
| "grad_norm": 0.33150729537010193, | |
| "learning_rate": 0.0005145637740244613, | |
| "loss": 3.418, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 7.159524918490917, | |
| "grad_norm": 0.3476800322532654, | |
| "learning_rate": 0.0005143890506697728, | |
| "loss": 3.3958, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 7.174080111783884, | |
| "grad_norm": 0.33134421706199646, | |
| "learning_rate": 0.0005142143273150844, | |
| "loss": 3.4261, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 7.1886353050768514, | |
| "grad_norm": 0.3467247188091278, | |
| "learning_rate": 0.000514039603960396, | |
| "loss": 3.4016, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 7.2031904983698185, | |
| "grad_norm": 0.3497604727745056, | |
| "learning_rate": 0.0005138648806057075, | |
| "loss": 3.419, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 7.217745691662786, | |
| "grad_norm": 0.32202965021133423, | |
| "learning_rate": 0.0005136901572510192, | |
| "loss": 3.411, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 7.232300884955753, | |
| "grad_norm": 0.33623242378234863, | |
| "learning_rate": 0.0005135154338963307, | |
| "loss": 3.4072, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 7.246856078248719, | |
| "grad_norm": 0.31815212965011597, | |
| "learning_rate": 0.0005133407105416424, | |
| "loss": 3.4238, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 7.261411271541686, | |
| "grad_norm": 0.32730230689048767, | |
| "learning_rate": 0.0005131659871869539, | |
| "loss": 3.424, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 7.275966464834653, | |
| "grad_norm": 0.3323996365070343, | |
| "learning_rate": 0.0005129912638322656, | |
| "loss": 3.4119, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.275966464834653, | |
| "eval_accuracy": 0.36525711859950705, | |
| "eval_loss": 3.5936996936798096, | |
| "eval_runtime": 83.8943, | |
| "eval_samples_per_second": 198.476, | |
| "eval_steps_per_second": 12.408, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.29052165812762, | |
| "grad_norm": 0.3421192169189453, | |
| "learning_rate": 0.0005128165404775771, | |
| "loss": 3.4307, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 7.305076851420587, | |
| "grad_norm": 0.31770020723342896, | |
| "learning_rate": 0.0005126418171228888, | |
| "loss": 3.4296, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 7.319632044713554, | |
| "grad_norm": 0.3201829493045807, | |
| "learning_rate": 0.0005124670937682003, | |
| "loss": 3.4104, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 7.334187238006521, | |
| "grad_norm": 0.3376303017139435, | |
| "learning_rate": 0.000512292370413512, | |
| "loss": 3.4243, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 7.348742431299487, | |
| "grad_norm": 0.3303367495536804, | |
| "learning_rate": 0.0005121176470588235, | |
| "loss": 3.4281, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 7.363297624592454, | |
| "grad_norm": 0.35328593850135803, | |
| "learning_rate": 0.000511942923704135, | |
| "loss": 3.4137, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 7.3778528178854215, | |
| "grad_norm": 0.3237740099430084, | |
| "learning_rate": 0.0005117682003494467, | |
| "loss": 3.4368, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 7.3924080111783885, | |
| "grad_norm": 0.30643945932388306, | |
| "learning_rate": 0.0005115934769947583, | |
| "loss": 3.4421, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 7.406963204471356, | |
| "grad_norm": 0.33868706226348877, | |
| "learning_rate": 0.0005114187536400699, | |
| "loss": 3.4258, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 7.421518397764323, | |
| "grad_norm": 0.32217374444007874, | |
| "learning_rate": 0.0005112440302853814, | |
| "loss": 3.4335, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 7.436073591057289, | |
| "grad_norm": 0.3170250654220581, | |
| "learning_rate": 0.0005110693069306931, | |
| "loss": 3.4368, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 7.450628784350256, | |
| "grad_norm": 0.31455597281455994, | |
| "learning_rate": 0.0005108945835760046, | |
| "loss": 3.4412, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 7.465183977643223, | |
| "grad_norm": 0.35698962211608887, | |
| "learning_rate": 0.0005107198602213162, | |
| "loss": 3.4364, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 7.47973917093619, | |
| "grad_norm": 0.3114458918571472, | |
| "learning_rate": 0.0005105451368666278, | |
| "loss": 3.4359, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 7.494294364229157, | |
| "grad_norm": 0.3224164843559265, | |
| "learning_rate": 0.0005103704135119394, | |
| "loss": 3.4398, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 7.508849557522124, | |
| "grad_norm": 0.31808143854141235, | |
| "learning_rate": 0.000510195690157251, | |
| "loss": 3.4523, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 7.523404750815091, | |
| "grad_norm": 0.313334196805954, | |
| "learning_rate": 0.0005100209668025625, | |
| "loss": 3.4273, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 7.537959944108058, | |
| "grad_norm": 0.32492595911026, | |
| "learning_rate": 0.0005098462434478742, | |
| "loss": 3.4338, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 7.552515137401024, | |
| "grad_norm": 0.34585487842559814, | |
| "learning_rate": 0.0005096715200931858, | |
| "loss": 3.4524, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 7.5670703306939915, | |
| "grad_norm": 0.3242366313934326, | |
| "learning_rate": 0.0005094967967384974, | |
| "loss": 3.4391, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.5670703306939915, | |
| "eval_accuracy": 0.3660836881384744, | |
| "eval_loss": 3.5846948623657227, | |
| "eval_runtime": 83.8105, | |
| "eval_samples_per_second": 198.674, | |
| "eval_steps_per_second": 12.421, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.5816255239869585, | |
| "grad_norm": 0.3198903203010559, | |
| "learning_rate": 0.0005093220733838089, | |
| "loss": 3.4446, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 7.596180717279926, | |
| "grad_norm": 0.31278297305107117, | |
| "learning_rate": 0.0005091473500291205, | |
| "loss": 3.4362, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 7.610735910572893, | |
| "grad_norm": 0.32255205512046814, | |
| "learning_rate": 0.0005089726266744321, | |
| "loss": 3.4552, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 7.625291103865859, | |
| "grad_norm": 0.30308541655540466, | |
| "learning_rate": 0.0005087979033197437, | |
| "loss": 3.4468, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 7.639846297158826, | |
| "grad_norm": 0.33427974581718445, | |
| "learning_rate": 0.0005086231799650553, | |
| "loss": 3.4523, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 7.654401490451793, | |
| "grad_norm": 0.3267691433429718, | |
| "learning_rate": 0.0005084484566103669, | |
| "loss": 3.4385, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 7.66895668374476, | |
| "grad_norm": 0.30869439244270325, | |
| "learning_rate": 0.0005082737332556785, | |
| "loss": 3.4455, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 7.683511877037727, | |
| "grad_norm": 0.29977184534072876, | |
| "learning_rate": 0.00050809900990099, | |
| "loss": 3.4521, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 7.698067070330694, | |
| "grad_norm": 0.3404894173145294, | |
| "learning_rate": 0.0005079242865463016, | |
| "loss": 3.4444, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 7.712622263623661, | |
| "grad_norm": 0.318911075592041, | |
| "learning_rate": 0.0005077495631916133, | |
| "loss": 3.4444, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 7.727177456916628, | |
| "grad_norm": 0.3126552700996399, | |
| "learning_rate": 0.0005075748398369248, | |
| "loss": 3.4497, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 7.7417326502095944, | |
| "grad_norm": 0.34629765152931213, | |
| "learning_rate": 0.0005074001164822364, | |
| "loss": 3.4589, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 7.7562878435025615, | |
| "grad_norm": 0.33652567863464355, | |
| "learning_rate": 0.000507225393127548, | |
| "loss": 3.4582, | |
| "step": 26650 | |
| }, | |
| { | |
| "epoch": 7.770843036795529, | |
| "grad_norm": 0.3080112636089325, | |
| "learning_rate": 0.0005070506697728596, | |
| "loss": 3.4522, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 7.785398230088496, | |
| "grad_norm": 0.32944127917289734, | |
| "learning_rate": 0.0005068759464181711, | |
| "loss": 3.448, | |
| "step": 26750 | |
| }, | |
| { | |
| "epoch": 7.799953423381463, | |
| "grad_norm": 0.3473741412162781, | |
| "learning_rate": 0.0005067012230634828, | |
| "loss": 3.4605, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 7.81450861667443, | |
| "grad_norm": 0.32413631677627563, | |
| "learning_rate": 0.0005065264997087944, | |
| "loss": 3.4474, | |
| "step": 26850 | |
| }, | |
| { | |
| "epoch": 7.829063809967396, | |
| "grad_norm": 0.3428443670272827, | |
| "learning_rate": 0.0005063517763541059, | |
| "loss": 3.4509, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 7.843619003260363, | |
| "grad_norm": 0.3344133198261261, | |
| "learning_rate": 0.0005061770529994175, | |
| "loss": 3.4519, | |
| "step": 26950 | |
| }, | |
| { | |
| "epoch": 7.85817419655333, | |
| "grad_norm": 0.3379044532775879, | |
| "learning_rate": 0.0005060023296447291, | |
| "loss": 3.4647, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.85817419655333, | |
| "eval_accuracy": 0.36677004760208454, | |
| "eval_loss": 3.575622081756592, | |
| "eval_runtime": 83.76, | |
| "eval_samples_per_second": 198.794, | |
| "eval_steps_per_second": 12.428, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 7.872729389846297, | |
| "grad_norm": 0.32137036323547363, | |
| "learning_rate": 0.0005058276062900408, | |
| "loss": 3.4514, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 7.887284583139264, | |
| "grad_norm": 0.3298042118549347, | |
| "learning_rate": 0.0005056528829353523, | |
| "loss": 3.4521, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 7.901839776432231, | |
| "grad_norm": 0.3261767327785492, | |
| "learning_rate": 0.000505478159580664, | |
| "loss": 3.4496, | |
| "step": 27150 | |
| }, | |
| { | |
| "epoch": 7.916394969725198, | |
| "grad_norm": 0.3207096457481384, | |
| "learning_rate": 0.0005053034362259755, | |
| "loss": 3.461, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 7.930950163018165, | |
| "grad_norm": 0.34332790970802307, | |
| "learning_rate": 0.000505128712871287, | |
| "loss": 3.449, | |
| "step": 27250 | |
| }, | |
| { | |
| "epoch": 7.9455053563111315, | |
| "grad_norm": 0.32024627923965454, | |
| "learning_rate": 0.0005049539895165987, | |
| "loss": 3.45, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 7.960060549604099, | |
| "grad_norm": 0.3165992796421051, | |
| "learning_rate": 0.0005047792661619103, | |
| "loss": 3.4513, | |
| "step": 27350 | |
| }, | |
| { | |
| "epoch": 7.974615742897066, | |
| "grad_norm": 0.312707781791687, | |
| "learning_rate": 0.0005046045428072219, | |
| "loss": 3.454, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 7.989170936190033, | |
| "grad_norm": 0.30599111318588257, | |
| "learning_rate": 0.0005044298194525334, | |
| "loss": 3.4538, | |
| "step": 27450 | |
| }, | |
| { | |
| "epoch": 8.003493246390311, | |
| "grad_norm": 0.34355446696281433, | |
| "learning_rate": 0.0005042550960978451, | |
| "loss": 3.4238, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 8.018048439683279, | |
| "grad_norm": 0.3354220390319824, | |
| "learning_rate": 0.0005040803727431566, | |
| "loss": 3.352, | |
| "step": 27550 | |
| }, | |
| { | |
| "epoch": 8.032603632976246, | |
| "grad_norm": 0.32905808091163635, | |
| "learning_rate": 0.0005039056493884683, | |
| "loss": 3.3508, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 8.047158826269213, | |
| "grad_norm": 0.3301757574081421, | |
| "learning_rate": 0.0005037309260337798, | |
| "loss": 3.3523, | |
| "step": 27650 | |
| }, | |
| { | |
| "epoch": 8.06171401956218, | |
| "grad_norm": 0.3272057771682739, | |
| "learning_rate": 0.0005035562026790914, | |
| "loss": 3.3572, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 8.076269212855147, | |
| "grad_norm": 0.31424692273139954, | |
| "learning_rate": 0.000503381479324403, | |
| "loss": 3.3527, | |
| "step": 27750 | |
| }, | |
| { | |
| "epoch": 8.090824406148114, | |
| "grad_norm": 0.337464839220047, | |
| "learning_rate": 0.0005032067559697145, | |
| "loss": 3.3793, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 8.10537959944108, | |
| "grad_norm": 0.35570797324180603, | |
| "learning_rate": 0.0005030320326150262, | |
| "loss": 3.3744, | |
| "step": 27850 | |
| }, | |
| { | |
| "epoch": 8.119934792734048, | |
| "grad_norm": 0.3274856209754944, | |
| "learning_rate": 0.0005028573092603378, | |
| "loss": 3.3817, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 8.134489986027015, | |
| "grad_norm": 0.32050132751464844, | |
| "learning_rate": 0.0005026825859056494, | |
| "loss": 3.377, | |
| "step": 27950 | |
| }, | |
| { | |
| "epoch": 8.149045179319982, | |
| "grad_norm": 0.33790844678878784, | |
| "learning_rate": 0.0005025078625509609, | |
| "loss": 3.3693, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.149045179319982, | |
| "eval_accuracy": 0.36650984214790083, | |
| "eval_loss": 3.58828067779541, | |
| "eval_runtime": 83.7404, | |
| "eval_samples_per_second": 198.841, | |
| "eval_steps_per_second": 12.431, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.16360037261295, | |
| "grad_norm": 0.3301652669906616, | |
| "learning_rate": 0.0005023331391962726, | |
| "loss": 3.3791, | |
| "step": 28050 | |
| }, | |
| { | |
| "epoch": 8.178155565905914, | |
| "grad_norm": 0.31026899814605713, | |
| "learning_rate": 0.0005021584158415841, | |
| "loss": 3.3704, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 8.192710759198881, | |
| "grad_norm": 0.3381020128726959, | |
| "learning_rate": 0.0005019836924868956, | |
| "loss": 3.3845, | |
| "step": 28150 | |
| }, | |
| { | |
| "epoch": 8.207265952491849, | |
| "grad_norm": 0.34272143244743347, | |
| "learning_rate": 0.0005018089691322073, | |
| "loss": 3.3783, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 8.221821145784816, | |
| "grad_norm": 0.3271588087081909, | |
| "learning_rate": 0.0005016342457775189, | |
| "loss": 3.3786, | |
| "step": 28250 | |
| }, | |
| { | |
| "epoch": 8.236376339077783, | |
| "grad_norm": 0.34381282329559326, | |
| "learning_rate": 0.0005014595224228305, | |
| "loss": 3.3773, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 8.25093153237075, | |
| "grad_norm": 0.3206559717655182, | |
| "learning_rate": 0.000501284799068142, | |
| "loss": 3.3938, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 8.265486725663717, | |
| "grad_norm": 0.33253544569015503, | |
| "learning_rate": 0.0005011100757134537, | |
| "loss": 3.3895, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 8.280041918956684, | |
| "grad_norm": 0.3516266942024231, | |
| "learning_rate": 0.0005009353523587653, | |
| "loss": 3.3976, | |
| "step": 28450 | |
| }, | |
| { | |
| "epoch": 8.294597112249651, | |
| "grad_norm": 0.3551691174507141, | |
| "learning_rate": 0.0005007606290040768, | |
| "loss": 3.3971, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 8.309152305542618, | |
| "grad_norm": 0.31077900528907776, | |
| "learning_rate": 0.0005005859056493884, | |
| "loss": 3.3994, | |
| "step": 28550 | |
| }, | |
| { | |
| "epoch": 8.323707498835585, | |
| "grad_norm": 0.3342590630054474, | |
| "learning_rate": 0.0005004111822947, | |
| "loss": 3.3956, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 8.338262692128552, | |
| "grad_norm": 0.3277079164981842, | |
| "learning_rate": 0.0005002364589400116, | |
| "loss": 3.4065, | |
| "step": 28650 | |
| }, | |
| { | |
| "epoch": 8.35281788542152, | |
| "grad_norm": 0.3241378366947174, | |
| "learning_rate": 0.0005000617355853231, | |
| "loss": 3.4006, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 8.367373078714486, | |
| "grad_norm": 0.3453556299209595, | |
| "learning_rate": 0.0004998870122306348, | |
| "loss": 3.3993, | |
| "step": 28750 | |
| }, | |
| { | |
| "epoch": 8.381928272007451, | |
| "grad_norm": 0.32487723231315613, | |
| "learning_rate": 0.0004997122888759464, | |
| "loss": 3.3952, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 8.396483465300419, | |
| "grad_norm": 0.3383024036884308, | |
| "learning_rate": 0.000499537565521258, | |
| "loss": 3.395, | |
| "step": 28850 | |
| }, | |
| { | |
| "epoch": 8.411038658593386, | |
| "grad_norm": 0.31621143221855164, | |
| "learning_rate": 0.0004993628421665695, | |
| "loss": 3.4038, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 8.425593851886353, | |
| "grad_norm": 0.3299100399017334, | |
| "learning_rate": 0.0004991881188118811, | |
| "loss": 3.4047, | |
| "step": 28950 | |
| }, | |
| { | |
| "epoch": 8.44014904517932, | |
| "grad_norm": 0.3217173218727112, | |
| "learning_rate": 0.0004990133954571928, | |
| "loss": 3.4013, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.44014904517932, | |
| "eval_accuracy": 0.367013681706205, | |
| "eval_loss": 3.5807859897613525, | |
| "eval_runtime": 83.8138, | |
| "eval_samples_per_second": 198.667, | |
| "eval_steps_per_second": 12.42, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.454704238472287, | |
| "grad_norm": 0.3334009647369385, | |
| "learning_rate": 0.0004988386721025043, | |
| "loss": 3.4077, | |
| "step": 29050 | |
| }, | |
| { | |
| "epoch": 8.469259431765254, | |
| "grad_norm": 0.3372880220413208, | |
| "learning_rate": 0.0004986639487478159, | |
| "loss": 3.4177, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 8.483814625058221, | |
| "grad_norm": 0.3230380415916443, | |
| "learning_rate": 0.0004984892253931275, | |
| "loss": 3.4138, | |
| "step": 29150 | |
| }, | |
| { | |
| "epoch": 8.498369818351188, | |
| "grad_norm": 0.3285495638847351, | |
| "learning_rate": 0.0004983145020384391, | |
| "loss": 3.4116, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 8.512925011644155, | |
| "grad_norm": 0.3134545385837555, | |
| "learning_rate": 0.0004981397786837507, | |
| "loss": 3.3921, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 8.527480204937122, | |
| "grad_norm": 0.3648611903190613, | |
| "learning_rate": 0.0004979650553290622, | |
| "loss": 3.4095, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 8.54203539823009, | |
| "grad_norm": 0.33360880613327026, | |
| "learning_rate": 0.0004977903319743739, | |
| "loss": 3.4083, | |
| "step": 29350 | |
| }, | |
| { | |
| "epoch": 8.556590591523056, | |
| "grad_norm": 0.3139150142669678, | |
| "learning_rate": 0.0004976156086196854, | |
| "loss": 3.4115, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 8.571145784816022, | |
| "grad_norm": 0.3379873037338257, | |
| "learning_rate": 0.0004974408852649971, | |
| "loss": 3.4191, | |
| "step": 29450 | |
| }, | |
| { | |
| "epoch": 8.585700978108989, | |
| "grad_norm": 0.3188300132751465, | |
| "learning_rate": 0.0004972661619103086, | |
| "loss": 3.4295, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 8.600256171401956, | |
| "grad_norm": 0.3309311866760254, | |
| "learning_rate": 0.0004970914385556202, | |
| "loss": 3.4093, | |
| "step": 29550 | |
| }, | |
| { | |
| "epoch": 8.614811364694923, | |
| "grad_norm": 0.3204537630081177, | |
| "learning_rate": 0.0004969167152009318, | |
| "loss": 3.4156, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 8.62936655798789, | |
| "grad_norm": 0.33092939853668213, | |
| "learning_rate": 0.0004967419918462435, | |
| "loss": 3.4245, | |
| "step": 29650 | |
| }, | |
| { | |
| "epoch": 8.643921751280857, | |
| "grad_norm": 0.3170015215873718, | |
| "learning_rate": 0.000496567268491555, | |
| "loss": 3.4138, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 8.658476944573824, | |
| "grad_norm": 0.3812834918498993, | |
| "learning_rate": 0.0004963925451368665, | |
| "loss": 3.4296, | |
| "step": 29750 | |
| }, | |
| { | |
| "epoch": 8.673032137866791, | |
| "grad_norm": 0.3209960162639618, | |
| "learning_rate": 0.0004962178217821782, | |
| "loss": 3.4165, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 8.687587331159758, | |
| "grad_norm": 0.33114925026893616, | |
| "learning_rate": 0.0004960430984274898, | |
| "loss": 3.4149, | |
| "step": 29850 | |
| }, | |
| { | |
| "epoch": 8.702142524452725, | |
| "grad_norm": 0.3401464521884918, | |
| "learning_rate": 0.0004958683750728014, | |
| "loss": 3.4174, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 8.716697717745692, | |
| "grad_norm": 0.3494839072227478, | |
| "learning_rate": 0.0004956936517181129, | |
| "loss": 3.4167, | |
| "step": 29950 | |
| }, | |
| { | |
| "epoch": 8.73125291103866, | |
| "grad_norm": 0.3174310028553009, | |
| "learning_rate": 0.0004955189283634246, | |
| "loss": 3.431, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.73125291103866, | |
| "eval_accuracy": 0.3679951522337063, | |
| "eval_loss": 3.5708882808685303, | |
| "eval_runtime": 83.8263, | |
| "eval_samples_per_second": 198.637, | |
| "eval_steps_per_second": 12.419, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 8.745808104331626, | |
| "grad_norm": 0.3153737187385559, | |
| "learning_rate": 0.0004953442050087361, | |
| "loss": 3.4291, | |
| "step": 30050 | |
| }, | |
| { | |
| "epoch": 8.760363297624593, | |
| "grad_norm": 0.3325824439525604, | |
| "learning_rate": 0.0004951694816540476, | |
| "loss": 3.4061, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 8.774918490917559, | |
| "grad_norm": 0.3704846501350403, | |
| "learning_rate": 0.0004949947582993593, | |
| "loss": 3.4194, | |
| "step": 30150 | |
| }, | |
| { | |
| "epoch": 8.789473684210526, | |
| "grad_norm": 0.3138867914676666, | |
| "learning_rate": 0.0004948200349446709, | |
| "loss": 3.4206, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 8.804028877503493, | |
| "grad_norm": 0.3463880121707916, | |
| "learning_rate": 0.0004946453115899825, | |
| "loss": 3.4287, | |
| "step": 30250 | |
| }, | |
| { | |
| "epoch": 8.81858407079646, | |
| "grad_norm": 0.3118993937969208, | |
| "learning_rate": 0.000494470588235294, | |
| "loss": 3.4282, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 8.833139264089427, | |
| "grad_norm": 0.3232675790786743, | |
| "learning_rate": 0.0004942958648806057, | |
| "loss": 3.427, | |
| "step": 30350 | |
| }, | |
| { | |
| "epoch": 8.847694457382394, | |
| "grad_norm": 0.33890658617019653, | |
| "learning_rate": 0.0004941211415259173, | |
| "loss": 3.42, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 8.862249650675361, | |
| "grad_norm": 0.3363267183303833, | |
| "learning_rate": 0.0004939464181712289, | |
| "loss": 3.4336, | |
| "step": 30450 | |
| }, | |
| { | |
| "epoch": 8.876804843968328, | |
| "grad_norm": 0.32357585430145264, | |
| "learning_rate": 0.0004937716948165404, | |
| "loss": 3.4363, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 8.891360037261295, | |
| "grad_norm": 0.30946317315101624, | |
| "learning_rate": 0.000493596971461852, | |
| "loss": 3.4357, | |
| "step": 30550 | |
| }, | |
| { | |
| "epoch": 8.905915230554262, | |
| "grad_norm": 0.32091355323791504, | |
| "learning_rate": 0.0004934222481071636, | |
| "loss": 3.4237, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 8.92047042384723, | |
| "grad_norm": 0.3205897808074951, | |
| "learning_rate": 0.0004932475247524751, | |
| "loss": 3.4112, | |
| "step": 30650 | |
| }, | |
| { | |
| "epoch": 8.935025617140196, | |
| "grad_norm": 0.3215363621711731, | |
| "learning_rate": 0.0004930728013977868, | |
| "loss": 3.4145, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 8.949580810433163, | |
| "grad_norm": 0.30593037605285645, | |
| "learning_rate": 0.0004928980780430984, | |
| "loss": 3.4343, | |
| "step": 30750 | |
| }, | |
| { | |
| "epoch": 8.964136003726129, | |
| "grad_norm": 0.3243584930896759, | |
| "learning_rate": 0.00049272335468841, | |
| "loss": 3.4221, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 8.978691197019096, | |
| "grad_norm": 0.3154670000076294, | |
| "learning_rate": 0.0004925486313337215, | |
| "loss": 3.4254, | |
| "step": 30850 | |
| }, | |
| { | |
| "epoch": 8.993246390312063, | |
| "grad_norm": 0.33407697081565857, | |
| "learning_rate": 0.0004923739079790332, | |
| "loss": 3.4227, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 9.007568700512342, | |
| "grad_norm": 0.3370547890663147, | |
| "learning_rate": 0.0004921991846243447, | |
| "loss": 3.3666, | |
| "step": 30950 | |
| }, | |
| { | |
| "epoch": 9.02212389380531, | |
| "grad_norm": 0.3225894570350647, | |
| "learning_rate": 0.0004920244612696563, | |
| "loss": 3.3154, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.02212389380531, | |
| "eval_accuracy": 0.3678739815818259, | |
| "eval_loss": 3.573394536972046, | |
| "eval_runtime": 83.8166, | |
| "eval_samples_per_second": 198.66, | |
| "eval_steps_per_second": 12.42, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.036679087098276, | |
| "grad_norm": 0.3302759528160095, | |
| "learning_rate": 0.0004918497379149679, | |
| "loss": 3.3282, | |
| "step": 31050 | |
| }, | |
| { | |
| "epoch": 9.051234280391244, | |
| "grad_norm": 0.32008126378059387, | |
| "learning_rate": 0.0004916750145602795, | |
| "loss": 3.3333, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 9.06578947368421, | |
| "grad_norm": 0.3511596918106079, | |
| "learning_rate": 0.0004915002912055911, | |
| "loss": 3.3283, | |
| "step": 31150 | |
| }, | |
| { | |
| "epoch": 9.080344666977178, | |
| "grad_norm": 0.33976656198501587, | |
| "learning_rate": 0.0004913255678509026, | |
| "loss": 3.3419, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 9.094899860270145, | |
| "grad_norm": 0.328326940536499, | |
| "learning_rate": 0.0004911508444962143, | |
| "loss": 3.3234, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 9.109455053563112, | |
| "grad_norm": 0.32798898220062256, | |
| "learning_rate": 0.0004909761211415259, | |
| "loss": 3.3389, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 9.124010246856079, | |
| "grad_norm": 0.3396635949611664, | |
| "learning_rate": 0.0004908013977868375, | |
| "loss": 3.347, | |
| "step": 31350 | |
| }, | |
| { | |
| "epoch": 9.138565440149046, | |
| "grad_norm": 0.36493322253227234, | |
| "learning_rate": 0.0004906266744321491, | |
| "loss": 3.3422, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 9.153120633442011, | |
| "grad_norm": 0.3240058720111847, | |
| "learning_rate": 0.0004904519510774606, | |
| "loss": 3.3427, | |
| "step": 31450 | |
| }, | |
| { | |
| "epoch": 9.167675826734978, | |
| "grad_norm": 0.3347022533416748, | |
| "learning_rate": 0.0004902772277227722, | |
| "loss": 3.3562, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 9.182231020027945, | |
| "grad_norm": 0.32579469680786133, | |
| "learning_rate": 0.0004901025043680838, | |
| "loss": 3.3665, | |
| "step": 31550 | |
| }, | |
| { | |
| "epoch": 9.196786213320912, | |
| "grad_norm": 0.3210705518722534, | |
| "learning_rate": 0.0004899277810133955, | |
| "loss": 3.3511, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 9.21134140661388, | |
| "grad_norm": 0.3276688754558563, | |
| "learning_rate": 0.000489753057658707, | |
| "loss": 3.3498, | |
| "step": 31650 | |
| }, | |
| { | |
| "epoch": 9.225896599906847, | |
| "grad_norm": 0.3511107563972473, | |
| "learning_rate": 0.0004895783343040186, | |
| "loss": 3.3508, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 9.240451793199814, | |
| "grad_norm": 0.3530363440513611, | |
| "learning_rate": 0.0004894036109493302, | |
| "loss": 3.3684, | |
| "step": 31750 | |
| }, | |
| { | |
| "epoch": 9.25500698649278, | |
| "grad_norm": 0.37356847524642944, | |
| "learning_rate": 0.0004892288875946419, | |
| "loss": 3.364, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 9.269562179785748, | |
| "grad_norm": 0.35008788108825684, | |
| "learning_rate": 0.0004890541642399534, | |
| "loss": 3.3621, | |
| "step": 31850 | |
| }, | |
| { | |
| "epoch": 9.284117373078715, | |
| "grad_norm": 0.31765300035476685, | |
| "learning_rate": 0.0004888794408852649, | |
| "loss": 3.3653, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 9.298672566371682, | |
| "grad_norm": 0.3152081370353699, | |
| "learning_rate": 0.0004887047175305766, | |
| "loss": 3.3651, | |
| "step": 31950 | |
| }, | |
| { | |
| "epoch": 9.313227759664649, | |
| "grad_norm": 0.3326716125011444, | |
| "learning_rate": 0.0004885299941758881, | |
| "loss": 3.3683, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.313227759664649, | |
| "eval_accuracy": 0.36780793123618394, | |
| "eval_loss": 3.5766756534576416, | |
| "eval_runtime": 83.9227, | |
| "eval_samples_per_second": 198.409, | |
| "eval_steps_per_second": 12.404, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.327782952957616, | |
| "grad_norm": 0.3099070191383362, | |
| "learning_rate": 0.0004883552708211997, | |
| "loss": 3.3616, | |
| "step": 32050 | |
| }, | |
| { | |
| "epoch": 9.342338146250583, | |
| "grad_norm": 0.3084421157836914, | |
| "learning_rate": 0.00048818054746651137, | |
| "loss": 3.3732, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 9.356893339543548, | |
| "grad_norm": 0.3281128406524658, | |
| "learning_rate": 0.0004880058241118229, | |
| "loss": 3.3739, | |
| "step": 32150 | |
| }, | |
| { | |
| "epoch": 9.371448532836515, | |
| "grad_norm": 0.3311002552509308, | |
| "learning_rate": 0.0004878311007571345, | |
| "loss": 3.3717, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 9.386003726129482, | |
| "grad_norm": 0.3228418231010437, | |
| "learning_rate": 0.0004876563774024461, | |
| "loss": 3.3788, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 9.40055891942245, | |
| "grad_norm": 0.3288918435573578, | |
| "learning_rate": 0.00048748165404775763, | |
| "loss": 3.3826, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 9.415114112715417, | |
| "grad_norm": 0.3360990881919861, | |
| "learning_rate": 0.0004873069306930693, | |
| "loss": 3.3797, | |
| "step": 32350 | |
| }, | |
| { | |
| "epoch": 9.429669306008384, | |
| "grad_norm": 0.3107812702655792, | |
| "learning_rate": 0.0004871322073383809, | |
| "loss": 3.368, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 9.44422449930135, | |
| "grad_norm": 0.3334508538246155, | |
| "learning_rate": 0.00048695748398369247, | |
| "loss": 3.376, | |
| "step": 32450 | |
| }, | |
| { | |
| "epoch": 9.458779692594318, | |
| "grad_norm": 0.3216943144798279, | |
| "learning_rate": 0.000486782760629004, | |
| "loss": 3.3941, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 9.473334885887285, | |
| "grad_norm": 0.33119189739227295, | |
| "learning_rate": 0.0004866080372743156, | |
| "loss": 3.3824, | |
| "step": 32550 | |
| }, | |
| { | |
| "epoch": 9.487890079180252, | |
| "grad_norm": 0.34190335869789124, | |
| "learning_rate": 0.0004864333139196272, | |
| "loss": 3.3786, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 9.502445272473219, | |
| "grad_norm": 0.3494519889354706, | |
| "learning_rate": 0.00048625859056493885, | |
| "loss": 3.3753, | |
| "step": 32650 | |
| }, | |
| { | |
| "epoch": 9.517000465766186, | |
| "grad_norm": 0.32106876373291016, | |
| "learning_rate": 0.0004860838672102504, | |
| "loss": 3.3933, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 9.531555659059153, | |
| "grad_norm": 0.31805312633514404, | |
| "learning_rate": 0.000485909143855562, | |
| "loss": 3.4052, | |
| "step": 32750 | |
| }, | |
| { | |
| "epoch": 9.546110852352118, | |
| "grad_norm": 0.3400896489620209, | |
| "learning_rate": 0.0004857344205008736, | |
| "loss": 3.3941, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 9.560666045645085, | |
| "grad_norm": 0.35194507241249084, | |
| "learning_rate": 0.00048555969714618517, | |
| "loss": 3.392, | |
| "step": 32850 | |
| }, | |
| { | |
| "epoch": 9.575221238938052, | |
| "grad_norm": 0.3494252860546112, | |
| "learning_rate": 0.0004853849737914967, | |
| "loss": 3.3751, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 9.58977643223102, | |
| "grad_norm": 0.3131227493286133, | |
| "learning_rate": 0.00048521025043680836, | |
| "loss": 3.3895, | |
| "step": 32950 | |
| }, | |
| { | |
| "epoch": 9.604331625523987, | |
| "grad_norm": 0.3253318965435028, | |
| "learning_rate": 0.00048503552708211995, | |
| "loss": 3.3906, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.604331625523987, | |
| "eval_accuracy": 0.3687079553410343, | |
| "eval_loss": 3.564946174621582, | |
| "eval_runtime": 83.8007, | |
| "eval_samples_per_second": 198.698, | |
| "eval_steps_per_second": 12.422, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.618886818816954, | |
| "grad_norm": 0.33878374099731445, | |
| "learning_rate": 0.00048486080372743155, | |
| "loss": 3.3888, | |
| "step": 33050 | |
| }, | |
| { | |
| "epoch": 9.63344201210992, | |
| "grad_norm": 0.3559328019618988, | |
| "learning_rate": 0.0004846860803727431, | |
| "loss": 3.4031, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 9.647997205402888, | |
| "grad_norm": 0.3159598410129547, | |
| "learning_rate": 0.0004845113570180547, | |
| "loss": 3.3861, | |
| "step": 33150 | |
| }, | |
| { | |
| "epoch": 9.662552398695855, | |
| "grad_norm": 0.3149755895137787, | |
| "learning_rate": 0.00048433663366336633, | |
| "loss": 3.391, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 9.677107591988822, | |
| "grad_norm": 0.34177547693252563, | |
| "learning_rate": 0.0004841619103086779, | |
| "loss": 3.4033, | |
| "step": 33250 | |
| }, | |
| { | |
| "epoch": 9.691662785281789, | |
| "grad_norm": 0.34913524985313416, | |
| "learning_rate": 0.00048398718695398947, | |
| "loss": 3.397, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 9.706217978574756, | |
| "grad_norm": 0.3287607431411743, | |
| "learning_rate": 0.00048381246359930106, | |
| "loss": 3.3985, | |
| "step": 33350 | |
| }, | |
| { | |
| "epoch": 9.720773171867723, | |
| "grad_norm": 0.35174891352653503, | |
| "learning_rate": 0.00048363774024461265, | |
| "loss": 3.3886, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 9.73532836516069, | |
| "grad_norm": 0.3212623596191406, | |
| "learning_rate": 0.0004834630168899242, | |
| "loss": 3.3967, | |
| "step": 33450 | |
| }, | |
| { | |
| "epoch": 9.749883558453657, | |
| "grad_norm": 0.3489495515823364, | |
| "learning_rate": 0.00048328829353523584, | |
| "loss": 3.3936, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 9.764438751746622, | |
| "grad_norm": 0.3314400613307953, | |
| "learning_rate": 0.00048311357018054744, | |
| "loss": 3.4043, | |
| "step": 33550 | |
| }, | |
| { | |
| "epoch": 9.77899394503959, | |
| "grad_norm": 0.3268505930900574, | |
| "learning_rate": 0.00048293884682585903, | |
| "loss": 3.3917, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 9.793549138332557, | |
| "grad_norm": 0.3297256529331207, | |
| "learning_rate": 0.00048276412347117057, | |
| "loss": 3.4011, | |
| "step": 33650 | |
| }, | |
| { | |
| "epoch": 9.808104331625524, | |
| "grad_norm": 0.3432904779911041, | |
| "learning_rate": 0.00048258940011648217, | |
| "loss": 3.4103, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 9.82265952491849, | |
| "grad_norm": 0.3400316536426544, | |
| "learning_rate": 0.0004824146767617938, | |
| "loss": 3.3996, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 9.837214718211458, | |
| "grad_norm": 0.34545719623565674, | |
| "learning_rate": 0.0004822399534071054, | |
| "loss": 3.395, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 9.851769911504425, | |
| "grad_norm": 0.3163986802101135, | |
| "learning_rate": 0.00048206523005241695, | |
| "loss": 3.4168, | |
| "step": 33850 | |
| }, | |
| { | |
| "epoch": 9.866325104797392, | |
| "grad_norm": 0.34051278233528137, | |
| "learning_rate": 0.00048189050669772854, | |
| "loss": 3.4083, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 9.880880298090359, | |
| "grad_norm": 0.3436630368232727, | |
| "learning_rate": 0.00048171578334304014, | |
| "loss": 3.4083, | |
| "step": 33950 | |
| }, | |
| { | |
| "epoch": 9.895435491383326, | |
| "grad_norm": 0.33097362518310547, | |
| "learning_rate": 0.00048154105998835173, | |
| "loss": 3.4058, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.895435491383326, | |
| "eval_accuracy": 0.3693286170409187, | |
| "eval_loss": 3.556833505630493, | |
| "eval_runtime": 83.802, | |
| "eval_samples_per_second": 198.695, | |
| "eval_steps_per_second": 12.422, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 9.909990684676293, | |
| "grad_norm": 0.3343755304813385, | |
| "learning_rate": 0.0004813663366336633, | |
| "loss": 3.4, | |
| "step": 34050 | |
| }, | |
| { | |
| "epoch": 9.92454587796926, | |
| "grad_norm": 0.3136387765407562, | |
| "learning_rate": 0.0004811916132789749, | |
| "loss": 3.4041, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 9.939101071262227, | |
| "grad_norm": 0.3382185101509094, | |
| "learning_rate": 0.0004810168899242865, | |
| "loss": 3.3944, | |
| "step": 34150 | |
| }, | |
| { | |
| "epoch": 9.953656264555192, | |
| "grad_norm": 0.3438982665538788, | |
| "learning_rate": 0.0004808421665695981, | |
| "loss": 3.4137, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 9.96821145784816, | |
| "grad_norm": 0.35343900322914124, | |
| "learning_rate": 0.00048066744321490965, | |
| "loss": 3.4108, | |
| "step": 34250 | |
| }, | |
| { | |
| "epoch": 9.982766651141127, | |
| "grad_norm": 0.334072470664978, | |
| "learning_rate": 0.00048049271986022124, | |
| "loss": 3.3913, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 9.997321844434094, | |
| "grad_norm": 0.3209112882614136, | |
| "learning_rate": 0.0004803179965055329, | |
| "loss": 3.4101, | |
| "step": 34350 | |
| }, | |
| { | |
| "epoch": 10.011644154634373, | |
| "grad_norm": 0.3629921078681946, | |
| "learning_rate": 0.0004801432731508445, | |
| "loss": 3.3216, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 10.02619934792734, | |
| "grad_norm": 0.31611087918281555, | |
| "learning_rate": 0.000479968549796156, | |
| "loss": 3.2979, | |
| "step": 34450 | |
| }, | |
| { | |
| "epoch": 10.040754541220307, | |
| "grad_norm": 0.3309492766857147, | |
| "learning_rate": 0.0004797938264414676, | |
| "loss": 3.2988, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 10.055309734513274, | |
| "grad_norm": 0.32886791229248047, | |
| "learning_rate": 0.0004796191030867792, | |
| "loss": 3.3036, | |
| "step": 34550 | |
| }, | |
| { | |
| "epoch": 10.069864927806242, | |
| "grad_norm": 0.34782519936561584, | |
| "learning_rate": 0.00047944437973209086, | |
| "loss": 3.3063, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 10.084420121099209, | |
| "grad_norm": 0.3429340422153473, | |
| "learning_rate": 0.0004792696563774024, | |
| "loss": 3.3085, | |
| "step": 34650 | |
| }, | |
| { | |
| "epoch": 10.098975314392176, | |
| "grad_norm": 0.3316565752029419, | |
| "learning_rate": 0.000479094933022714, | |
| "loss": 3.3186, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 10.113530507685143, | |
| "grad_norm": 0.34856709837913513, | |
| "learning_rate": 0.0004789202096680256, | |
| "loss": 3.3198, | |
| "step": 34750 | |
| }, | |
| { | |
| "epoch": 10.12808570097811, | |
| "grad_norm": 0.3699038028717041, | |
| "learning_rate": 0.00047874548631333713, | |
| "loss": 3.3131, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 10.142640894271075, | |
| "grad_norm": 0.3273329734802246, | |
| "learning_rate": 0.0004785707629586487, | |
| "loss": 3.3266, | |
| "step": 34850 | |
| }, | |
| { | |
| "epoch": 10.157196087564042, | |
| "grad_norm": 0.3579559028148651, | |
| "learning_rate": 0.0004783960396039604, | |
| "loss": 3.3187, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 10.17175128085701, | |
| "grad_norm": 0.32014134526252747, | |
| "learning_rate": 0.00047822131624927197, | |
| "loss": 3.3342, | |
| "step": 34950 | |
| }, | |
| { | |
| "epoch": 10.186306474149976, | |
| "grad_norm": 0.3189047574996948, | |
| "learning_rate": 0.0004780465928945835, | |
| "loss": 3.3313, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.186306474149976, | |
| "eval_accuracy": 0.3687564941181697, | |
| "eval_loss": 3.5707647800445557, | |
| "eval_runtime": 83.7624, | |
| "eval_samples_per_second": 198.789, | |
| "eval_steps_per_second": 12.428, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.200861667442943, | |
| "grad_norm": 0.33648446202278137, | |
| "learning_rate": 0.0004778718695398951, | |
| "loss": 3.3374, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 10.21541686073591, | |
| "grad_norm": 0.3168446123600006, | |
| "learning_rate": 0.0004776971461852067, | |
| "loss": 3.3384, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 10.229972054028877, | |
| "grad_norm": 0.3517572581768036, | |
| "learning_rate": 0.00047752242283051835, | |
| "loss": 3.3426, | |
| "step": 35150 | |
| }, | |
| { | |
| "epoch": 10.244527247321844, | |
| "grad_norm": 0.3478202819824219, | |
| "learning_rate": 0.00047734769947582994, | |
| "loss": 3.3392, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 10.259082440614812, | |
| "grad_norm": 0.3541315197944641, | |
| "learning_rate": 0.0004771729761211415, | |
| "loss": 3.3449, | |
| "step": 35250 | |
| }, | |
| { | |
| "epoch": 10.273637633907779, | |
| "grad_norm": 0.34404176473617554, | |
| "learning_rate": 0.0004769982527664531, | |
| "loss": 3.3383, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 10.288192827200746, | |
| "grad_norm": 0.34748905897140503, | |
| "learning_rate": 0.00047682352941176467, | |
| "loss": 3.3532, | |
| "step": 35350 | |
| }, | |
| { | |
| "epoch": 10.302748020493713, | |
| "grad_norm": 0.3371129333972931, | |
| "learning_rate": 0.0004766488060570762, | |
| "loss": 3.3371, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 10.31730321378668, | |
| "grad_norm": 0.3395603895187378, | |
| "learning_rate": 0.00047647408270238786, | |
| "loss": 3.3537, | |
| "step": 35450 | |
| }, | |
| { | |
| "epoch": 10.331858407079647, | |
| "grad_norm": 0.3195820748806, | |
| "learning_rate": 0.00047629935934769945, | |
| "loss": 3.3604, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 10.346413600372612, | |
| "grad_norm": 0.3302899897098541, | |
| "learning_rate": 0.00047612463599301105, | |
| "loss": 3.3432, | |
| "step": 35550 | |
| }, | |
| { | |
| "epoch": 10.36096879366558, | |
| "grad_norm": 0.34034964442253113, | |
| "learning_rate": 0.0004759499126383226, | |
| "loss": 3.3384, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 10.375523986958546, | |
| "grad_norm": 0.3405877947807312, | |
| "learning_rate": 0.0004757751892836342, | |
| "loss": 3.3577, | |
| "step": 35650 | |
| }, | |
| { | |
| "epoch": 10.390079180251513, | |
| "grad_norm": 0.3528953492641449, | |
| "learning_rate": 0.0004756004659289458, | |
| "loss": 3.3481, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 10.40463437354448, | |
| "grad_norm": 0.3363061249256134, | |
| "learning_rate": 0.0004754257425742574, | |
| "loss": 3.35, | |
| "step": 35750 | |
| }, | |
| { | |
| "epoch": 10.419189566837447, | |
| "grad_norm": 0.3389376997947693, | |
| "learning_rate": 0.00047525101921956896, | |
| "loss": 3.3569, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 10.433744760130415, | |
| "grad_norm": 0.34698837995529175, | |
| "learning_rate": 0.00047507629586488056, | |
| "loss": 3.3681, | |
| "step": 35850 | |
| }, | |
| { | |
| "epoch": 10.448299953423382, | |
| "grad_norm": 0.328369677066803, | |
| "learning_rate": 0.00047490157251019215, | |
| "loss": 3.3488, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 10.462855146716349, | |
| "grad_norm": 0.34849104285240173, | |
| "learning_rate": 0.0004747268491555037, | |
| "loss": 3.3682, | |
| "step": 35950 | |
| }, | |
| { | |
| "epoch": 10.477410340009316, | |
| "grad_norm": 0.30830472707748413, | |
| "learning_rate": 0.00047455212580081534, | |
| "loss": 3.356, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.477410340009316, | |
| "eval_accuracy": 0.3691290556763279, | |
| "eval_loss": 3.5604703426361084, | |
| "eval_runtime": 83.7848, | |
| "eval_samples_per_second": 198.735, | |
| "eval_steps_per_second": 12.425, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.491965533302283, | |
| "grad_norm": 0.32924532890319824, | |
| "learning_rate": 0.00047437740244612694, | |
| "loss": 3.3553, | |
| "step": 36050 | |
| }, | |
| { | |
| "epoch": 10.50652072659525, | |
| "grad_norm": 0.3294248580932617, | |
| "learning_rate": 0.00047420267909143853, | |
| "loss": 3.3749, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 10.521075919888217, | |
| "grad_norm": 0.3199954926967621, | |
| "learning_rate": 0.0004740279557367501, | |
| "loss": 3.3673, | |
| "step": 36150 | |
| }, | |
| { | |
| "epoch": 10.535631113181182, | |
| "grad_norm": 0.3224357068538666, | |
| "learning_rate": 0.00047385323238206166, | |
| "loss": 3.3642, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 10.55018630647415, | |
| "grad_norm": 0.31655916571617126, | |
| "learning_rate": 0.00047367850902737326, | |
| "loss": 3.3713, | |
| "step": 36250 | |
| }, | |
| { | |
| "epoch": 10.564741499767116, | |
| "grad_norm": 0.36199480295181274, | |
| "learning_rate": 0.0004735037856726849, | |
| "loss": 3.3585, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 10.579296693060083, | |
| "grad_norm": 0.3338967263698578, | |
| "learning_rate": 0.0004733290623179965, | |
| "loss": 3.3691, | |
| "step": 36350 | |
| }, | |
| { | |
| "epoch": 10.59385188635305, | |
| "grad_norm": 0.325834721326828, | |
| "learning_rate": 0.00047315433896330804, | |
| "loss": 3.3769, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 10.608407079646017, | |
| "grad_norm": 0.32152023911476135, | |
| "learning_rate": 0.00047297961560861964, | |
| "loss": 3.3782, | |
| "step": 36450 | |
| }, | |
| { | |
| "epoch": 10.622962272938985, | |
| "grad_norm": 0.32181116938591003, | |
| "learning_rate": 0.00047280489225393123, | |
| "loss": 3.3732, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 10.637517466231952, | |
| "grad_norm": 0.3409736752510071, | |
| "learning_rate": 0.0004726301688992429, | |
| "loss": 3.3754, | |
| "step": 36550 | |
| }, | |
| { | |
| "epoch": 10.652072659524919, | |
| "grad_norm": 0.3407434821128845, | |
| "learning_rate": 0.0004724554455445544, | |
| "loss": 3.3721, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 10.666627852817886, | |
| "grad_norm": 0.3224821090698242, | |
| "learning_rate": 0.000472280722189866, | |
| "loss": 3.363, | |
| "step": 36650 | |
| }, | |
| { | |
| "epoch": 10.681183046110853, | |
| "grad_norm": 0.3334067165851593, | |
| "learning_rate": 0.0004721059988351776, | |
| "loss": 3.3675, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 10.69573823940382, | |
| "grad_norm": 0.3512733280658722, | |
| "learning_rate": 0.00047193127548048915, | |
| "loss": 3.3712, | |
| "step": 36750 | |
| }, | |
| { | |
| "epoch": 10.710293432696787, | |
| "grad_norm": 0.3165178894996643, | |
| "learning_rate": 0.00047175655212580074, | |
| "loss": 3.3744, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 10.724848625989754, | |
| "grad_norm": 0.3261968195438385, | |
| "learning_rate": 0.0004715818287711124, | |
| "loss": 3.396, | |
| "step": 36850 | |
| }, | |
| { | |
| "epoch": 10.73940381928272, | |
| "grad_norm": 0.3375428318977356, | |
| "learning_rate": 0.000471407105416424, | |
| "loss": 3.3719, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 10.753959012575686, | |
| "grad_norm": 0.32842355966567993, | |
| "learning_rate": 0.0004712323820617355, | |
| "loss": 3.3763, | |
| "step": 36950 | |
| }, | |
| { | |
| "epoch": 10.768514205868653, | |
| "grad_norm": 0.34244853258132935, | |
| "learning_rate": 0.0004710576587070471, | |
| "loss": 3.3801, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.768514205868653, | |
| "eval_accuracy": 0.3699390538652321, | |
| "eval_loss": 3.5535242557525635, | |
| "eval_runtime": 83.8183, | |
| "eval_samples_per_second": 198.656, | |
| "eval_steps_per_second": 12.42, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 10.78306939916162, | |
| "grad_norm": 0.36360105872154236, | |
| "learning_rate": 0.0004708829353523587, | |
| "loss": 3.3804, | |
| "step": 37050 | |
| }, | |
| { | |
| "epoch": 10.797624592454587, | |
| "grad_norm": 0.335449755191803, | |
| "learning_rate": 0.0004707082119976703, | |
| "loss": 3.3747, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 10.812179785747555, | |
| "grad_norm": 0.33392956852912903, | |
| "learning_rate": 0.0004705334886429819, | |
| "loss": 3.3876, | |
| "step": 37150 | |
| }, | |
| { | |
| "epoch": 10.826734979040522, | |
| "grad_norm": 0.3063945472240448, | |
| "learning_rate": 0.0004703587652882935, | |
| "loss": 3.3692, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 10.841290172333489, | |
| "grad_norm": 0.35233259201049805, | |
| "learning_rate": 0.0004701840419336051, | |
| "loss": 3.3654, | |
| "step": 37250 | |
| }, | |
| { | |
| "epoch": 10.855845365626456, | |
| "grad_norm": 0.3215087950229645, | |
| "learning_rate": 0.0004700093185789167, | |
| "loss": 3.3805, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 10.870400558919423, | |
| "grad_norm": 0.3286418318748474, | |
| "learning_rate": 0.0004698345952242282, | |
| "loss": 3.3657, | |
| "step": 37350 | |
| }, | |
| { | |
| "epoch": 10.88495575221239, | |
| "grad_norm": 0.35888421535491943, | |
| "learning_rate": 0.00046965987186953987, | |
| "loss": 3.3831, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 10.899510945505357, | |
| "grad_norm": 0.3441259562969208, | |
| "learning_rate": 0.00046948514851485147, | |
| "loss": 3.3856, | |
| "step": 37450 | |
| }, | |
| { | |
| "epoch": 10.914066138798324, | |
| "grad_norm": 0.32109716534614563, | |
| "learning_rate": 0.00046931042516016306, | |
| "loss": 3.3886, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 10.92862133209129, | |
| "grad_norm": 0.30541130900382996, | |
| "learning_rate": 0.0004691357018054746, | |
| "loss": 3.3883, | |
| "step": 37550 | |
| }, | |
| { | |
| "epoch": 10.943176525384256, | |
| "grad_norm": 0.3508901000022888, | |
| "learning_rate": 0.0004689609784507862, | |
| "loss": 3.3814, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 10.957731718677223, | |
| "grad_norm": 0.32192423939704895, | |
| "learning_rate": 0.0004687862550960978, | |
| "loss": 3.383, | |
| "step": 37650 | |
| }, | |
| { | |
| "epoch": 10.97228691197019, | |
| "grad_norm": 0.3409295976161957, | |
| "learning_rate": 0.00046861153174140944, | |
| "loss": 3.3887, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 10.986842105263158, | |
| "grad_norm": 0.3365081548690796, | |
| "learning_rate": 0.000468436808386721, | |
| "loss": 3.3791, | |
| "step": 37750 | |
| }, | |
| { | |
| "epoch": 11.001164415463437, | |
| "grad_norm": 0.3764905035495758, | |
| "learning_rate": 0.0004682620850320326, | |
| "loss": 3.3695, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 11.015719608756404, | |
| "grad_norm": 0.37236666679382324, | |
| "learning_rate": 0.00046808736167734417, | |
| "loss": 3.2656, | |
| "step": 37850 | |
| }, | |
| { | |
| "epoch": 11.030274802049371, | |
| "grad_norm": 0.34664493799209595, | |
| "learning_rate": 0.0004679126383226557, | |
| "loss": 3.2603, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 11.044829995342338, | |
| "grad_norm": 0.34286394715309143, | |
| "learning_rate": 0.00046773791496796736, | |
| "loss": 3.2878, | |
| "step": 37950 | |
| }, | |
| { | |
| "epoch": 11.059385188635305, | |
| "grad_norm": 0.35410016775131226, | |
| "learning_rate": 0.00046756319161327895, | |
| "loss": 3.2818, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.059385188635305, | |
| "eval_accuracy": 0.36981941106832206, | |
| "eval_loss": 3.5628693103790283, | |
| "eval_runtime": 83.7739, | |
| "eval_samples_per_second": 198.761, | |
| "eval_steps_per_second": 12.426, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.073940381928272, | |
| "grad_norm": 0.36358770728111267, | |
| "learning_rate": 0.00046738846825859054, | |
| "loss": 3.2948, | |
| "step": 38050 | |
| }, | |
| { | |
| "epoch": 11.08849557522124, | |
| "grad_norm": 0.3897538185119629, | |
| "learning_rate": 0.0004672137449039021, | |
| "loss": 3.2881, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 11.103050768514207, | |
| "grad_norm": 0.32897424697875977, | |
| "learning_rate": 0.0004670390215492137, | |
| "loss": 3.2868, | |
| "step": 38150 | |
| }, | |
| { | |
| "epoch": 11.117605961807174, | |
| "grad_norm": 0.35240480303764343, | |
| "learning_rate": 0.0004668642981945253, | |
| "loss": 3.2893, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 11.132161155100139, | |
| "grad_norm": 0.36881744861602783, | |
| "learning_rate": 0.0004666895748398369, | |
| "loss": 3.3004, | |
| "step": 38250 | |
| }, | |
| { | |
| "epoch": 11.146716348393106, | |
| "grad_norm": 0.35900160670280457, | |
| "learning_rate": 0.00046651485148514846, | |
| "loss": 3.3037, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 11.161271541686073, | |
| "grad_norm": 0.3306093215942383, | |
| "learning_rate": 0.00046634012813046006, | |
| "loss": 3.3011, | |
| "step": 38350 | |
| }, | |
| { | |
| "epoch": 11.17582673497904, | |
| "grad_norm": 0.34579455852508545, | |
| "learning_rate": 0.00046616540477577165, | |
| "loss": 3.3002, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 11.190381928272007, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00046599068142108324, | |
| "loss": 3.3123, | |
| "step": 38450 | |
| }, | |
| { | |
| "epoch": 11.204937121564974, | |
| "grad_norm": 0.36285683512687683, | |
| "learning_rate": 0.0004658159580663948, | |
| "loss": 3.3104, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 11.219492314857941, | |
| "grad_norm": 0.3402998447418213, | |
| "learning_rate": 0.00046564123471170643, | |
| "loss": 3.3225, | |
| "step": 38550 | |
| }, | |
| { | |
| "epoch": 11.234047508150908, | |
| "grad_norm": 0.34194666147232056, | |
| "learning_rate": 0.00046546651135701803, | |
| "loss": 3.3202, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 11.248602701443875, | |
| "grad_norm": 0.32852575182914734, | |
| "learning_rate": 0.0004652917880023296, | |
| "loss": 3.3147, | |
| "step": 38650 | |
| }, | |
| { | |
| "epoch": 11.263157894736842, | |
| "grad_norm": 0.3370783030986786, | |
| "learning_rate": 0.00046511706464764116, | |
| "loss": 3.3146, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 11.27771308802981, | |
| "grad_norm": 0.3787709176540375, | |
| "learning_rate": 0.00046494234129295276, | |
| "loss": 3.3316, | |
| "step": 38750 | |
| }, | |
| { | |
| "epoch": 11.292268281322777, | |
| "grad_norm": 0.3444885015487671, | |
| "learning_rate": 0.0004647676179382644, | |
| "loss": 3.313, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 11.306823474615744, | |
| "grad_norm": 0.33554887771606445, | |
| "learning_rate": 0.000464592894583576, | |
| "loss": 3.3282, | |
| "step": 38850 | |
| }, | |
| { | |
| "epoch": 11.32137866790871, | |
| "grad_norm": 0.33529722690582275, | |
| "learning_rate": 0.00046441817122888754, | |
| "loss": 3.3375, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 11.335933861201676, | |
| "grad_norm": 0.34856700897216797, | |
| "learning_rate": 0.00046424344787419913, | |
| "loss": 3.3453, | |
| "step": 38950 | |
| }, | |
| { | |
| "epoch": 11.350489054494643, | |
| "grad_norm": 0.3685392737388611, | |
| "learning_rate": 0.00046406872451951073, | |
| "loss": 3.3199, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.350489054494643, | |
| "eval_accuracy": 0.36944790725591253, | |
| "eval_loss": 3.5641021728515625, | |
| "eval_runtime": 83.8403, | |
| "eval_samples_per_second": 198.604, | |
| "eval_steps_per_second": 12.416, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.36504424778761, | |
| "grad_norm": 0.34376809000968933, | |
| "learning_rate": 0.00046389400116482227, | |
| "loss": 3.3276, | |
| "step": 39050 | |
| }, | |
| { | |
| "epoch": 11.379599441080577, | |
| "grad_norm": 0.3262241780757904, | |
| "learning_rate": 0.0004637192778101339, | |
| "loss": 3.3405, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 11.394154634373544, | |
| "grad_norm": 0.33913594484329224, | |
| "learning_rate": 0.0004635445544554455, | |
| "loss": 3.3351, | |
| "step": 39150 | |
| }, | |
| { | |
| "epoch": 11.408709827666511, | |
| "grad_norm": 0.3396729528903961, | |
| "learning_rate": 0.0004633698311007571, | |
| "loss": 3.3358, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 11.423265020959478, | |
| "grad_norm": 0.3508078455924988, | |
| "learning_rate": 0.0004631951077460687, | |
| "loss": 3.3371, | |
| "step": 39250 | |
| }, | |
| { | |
| "epoch": 11.437820214252445, | |
| "grad_norm": 0.3164216876029968, | |
| "learning_rate": 0.00046302038439138024, | |
| "loss": 3.3431, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 11.452375407545413, | |
| "grad_norm": 0.328569233417511, | |
| "learning_rate": 0.0004628456610366919, | |
| "loss": 3.3503, | |
| "step": 39350 | |
| }, | |
| { | |
| "epoch": 11.46693060083838, | |
| "grad_norm": 0.3737082779407501, | |
| "learning_rate": 0.0004626709376820035, | |
| "loss": 3.337, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 11.481485794131347, | |
| "grad_norm": 0.3797398805618286, | |
| "learning_rate": 0.0004624962143273151, | |
| "loss": 3.3492, | |
| "step": 39450 | |
| }, | |
| { | |
| "epoch": 11.496040987424314, | |
| "grad_norm": 0.3469921052455902, | |
| "learning_rate": 0.0004623214909726266, | |
| "loss": 3.3395, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 11.51059618071728, | |
| "grad_norm": 0.3705849349498749, | |
| "learning_rate": 0.0004621467676179382, | |
| "loss": 3.3538, | |
| "step": 39550 | |
| }, | |
| { | |
| "epoch": 11.525151374010246, | |
| "grad_norm": 0.331779807806015, | |
| "learning_rate": 0.0004619720442632498, | |
| "loss": 3.3308, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 11.539706567303213, | |
| "grad_norm": 0.34028103947639465, | |
| "learning_rate": 0.00046179732090856145, | |
| "loss": 3.3493, | |
| "step": 39650 | |
| }, | |
| { | |
| "epoch": 11.55426176059618, | |
| "grad_norm": 0.3361406922340393, | |
| "learning_rate": 0.000461622597553873, | |
| "loss": 3.3571, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 11.568816953889147, | |
| "grad_norm": 0.3408490717411041, | |
| "learning_rate": 0.0004614478741991846, | |
| "loss": 3.3443, | |
| "step": 39750 | |
| }, | |
| { | |
| "epoch": 11.583372147182114, | |
| "grad_norm": 0.34149110317230225, | |
| "learning_rate": 0.0004612731508444962, | |
| "loss": 3.3466, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 11.597927340475081, | |
| "grad_norm": 0.34090861678123474, | |
| "learning_rate": 0.0004610984274898077, | |
| "loss": 3.3501, | |
| "step": 39850 | |
| }, | |
| { | |
| "epoch": 11.612482533768048, | |
| "grad_norm": 0.3454326391220093, | |
| "learning_rate": 0.00046092370413511937, | |
| "loss": 3.3459, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 11.627037727061015, | |
| "grad_norm": 0.37060949206352234, | |
| "learning_rate": 0.00046074898078043096, | |
| "loss": 3.3473, | |
| "step": 39950 | |
| }, | |
| { | |
| "epoch": 11.641592920353983, | |
| "grad_norm": 0.33535870909690857, | |
| "learning_rate": 0.00046057425742574256, | |
| "loss": 3.3573, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.641592920353983, | |
| "eval_accuracy": 0.3703923566822089, | |
| "eval_loss": 3.5545589923858643, | |
| "eval_runtime": 83.7489, | |
| "eval_samples_per_second": 198.821, | |
| "eval_steps_per_second": 12.43, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 11.65614811364695, | |
| "grad_norm": 0.3319244980812073, | |
| "learning_rate": 0.0004603995340710541, | |
| "loss": 3.3488, | |
| "step": 40050 | |
| }, | |
| { | |
| "epoch": 11.670703306939917, | |
| "grad_norm": 0.33975809812545776, | |
| "learning_rate": 0.0004602248107163657, | |
| "loss": 3.3547, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 11.685258500232884, | |
| "grad_norm": 0.3551541268825531, | |
| "learning_rate": 0.0004600500873616773, | |
| "loss": 3.3575, | |
| "step": 40150 | |
| }, | |
| { | |
| "epoch": 11.69981369352585, | |
| "grad_norm": 0.3273112177848816, | |
| "learning_rate": 0.00045987536400698894, | |
| "loss": 3.3543, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 11.714368886818818, | |
| "grad_norm": 0.3514944911003113, | |
| "learning_rate": 0.0004597006406523005, | |
| "loss": 3.3483, | |
| "step": 40250 | |
| }, | |
| { | |
| "epoch": 11.728924080111783, | |
| "grad_norm": 0.3633091151714325, | |
| "learning_rate": 0.00045952591729761207, | |
| "loss": 3.3523, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 11.74347927340475, | |
| "grad_norm": 0.34564444422721863, | |
| "learning_rate": 0.00045935119394292367, | |
| "loss": 3.3609, | |
| "step": 40350 | |
| }, | |
| { | |
| "epoch": 11.758034466697717, | |
| "grad_norm": 0.3390924334526062, | |
| "learning_rate": 0.00045917647058823526, | |
| "loss": 3.3592, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 11.772589659990684, | |
| "grad_norm": 0.3222717046737671, | |
| "learning_rate": 0.0004590017472335468, | |
| "loss": 3.3484, | |
| "step": 40450 | |
| }, | |
| { | |
| "epoch": 11.787144853283651, | |
| "grad_norm": 0.3520161211490631, | |
| "learning_rate": 0.00045882702387885845, | |
| "loss": 3.3613, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 11.801700046576618, | |
| "grad_norm": 0.34458187222480774, | |
| "learning_rate": 0.00045865230052417004, | |
| "loss": 3.3673, | |
| "step": 40550 | |
| }, | |
| { | |
| "epoch": 11.816255239869585, | |
| "grad_norm": 0.3838929235935211, | |
| "learning_rate": 0.00045847757716948164, | |
| "loss": 3.365, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 11.830810433162553, | |
| "grad_norm": 0.3497732877731323, | |
| "learning_rate": 0.0004583028538147932, | |
| "loss": 3.3653, | |
| "step": 40650 | |
| }, | |
| { | |
| "epoch": 11.84536562645552, | |
| "grad_norm": 0.34237682819366455, | |
| "learning_rate": 0.00045812813046010477, | |
| "loss": 3.3581, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 11.859920819748487, | |
| "grad_norm": 0.3361947238445282, | |
| "learning_rate": 0.0004579534071054164, | |
| "loss": 3.3535, | |
| "step": 40750 | |
| }, | |
| { | |
| "epoch": 11.874476013041454, | |
| "grad_norm": 0.3154624104499817, | |
| "learning_rate": 0.000457778683750728, | |
| "loss": 3.3647, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 11.88903120633442, | |
| "grad_norm": 0.3609897494316101, | |
| "learning_rate": 0.00045760396039603955, | |
| "loss": 3.3532, | |
| "step": 40850 | |
| }, | |
| { | |
| "epoch": 11.903586399627388, | |
| "grad_norm": 0.34533870220184326, | |
| "learning_rate": 0.00045742923704135115, | |
| "loss": 3.3712, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 11.918141592920353, | |
| "grad_norm": 0.34753355383872986, | |
| "learning_rate": 0.00045725451368666274, | |
| "loss": 3.3696, | |
| "step": 40950 | |
| }, | |
| { | |
| "epoch": 11.93269678621332, | |
| "grad_norm": 0.33941614627838135, | |
| "learning_rate": 0.0004570797903319743, | |
| "loss": 3.3593, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.93269678621332, | |
| "eval_accuracy": 0.37077690602551916, | |
| "eval_loss": 3.5441534519195557, | |
| "eval_runtime": 83.8403, | |
| "eval_samples_per_second": 198.604, | |
| "eval_steps_per_second": 12.416, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 11.947251979506287, | |
| "grad_norm": 0.35061389207839966, | |
| "learning_rate": 0.00045690506697728593, | |
| "loss": 3.3699, | |
| "step": 41050 | |
| }, | |
| { | |
| "epoch": 11.961807172799254, | |
| "grad_norm": 0.33409953117370605, | |
| "learning_rate": 0.0004567303436225975, | |
| "loss": 3.3651, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 11.976362366092221, | |
| "grad_norm": 0.3518408238887787, | |
| "learning_rate": 0.0004565556202679091, | |
| "loss": 3.3644, | |
| "step": 41150 | |
| }, | |
| { | |
| "epoch": 11.990917559385188, | |
| "grad_norm": 0.34396883845329285, | |
| "learning_rate": 0.00045638089691322066, | |
| "loss": 3.3767, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 12.005239869585468, | |
| "grad_norm": 0.36460235714912415, | |
| "learning_rate": 0.00045620617355853225, | |
| "loss": 3.3361, | |
| "step": 41250 | |
| }, | |
| { | |
| "epoch": 12.019795062878435, | |
| "grad_norm": 0.3513859808444977, | |
| "learning_rate": 0.0004560314502038439, | |
| "loss": 3.257, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 12.034350256171402, | |
| "grad_norm": 0.36926543712615967, | |
| "learning_rate": 0.0004558567268491555, | |
| "loss": 3.2589, | |
| "step": 41350 | |
| }, | |
| { | |
| "epoch": 12.04890544946437, | |
| "grad_norm": 0.3499617576599121, | |
| "learning_rate": 0.00045568200349446704, | |
| "loss": 3.2655, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 12.063460642757336, | |
| "grad_norm": 0.3285706341266632, | |
| "learning_rate": 0.00045550728013977863, | |
| "loss": 3.2641, | |
| "step": 41450 | |
| }, | |
| { | |
| "epoch": 12.078015836050303, | |
| "grad_norm": 0.33528247475624084, | |
| "learning_rate": 0.0004553325567850902, | |
| "loss": 3.2581, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 12.09257102934327, | |
| "grad_norm": 0.34470561146736145, | |
| "learning_rate": 0.0004551578334304018, | |
| "loss": 3.282, | |
| "step": 41550 | |
| }, | |
| { | |
| "epoch": 12.107126222636236, | |
| "grad_norm": 0.3587476909160614, | |
| "learning_rate": 0.00045498311007571347, | |
| "loss": 3.2604, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 12.121681415929203, | |
| "grad_norm": 0.3634476363658905, | |
| "learning_rate": 0.000454808386721025, | |
| "loss": 3.2796, | |
| "step": 41650 | |
| }, | |
| { | |
| "epoch": 12.13623660922217, | |
| "grad_norm": 0.4091065227985382, | |
| "learning_rate": 0.0004546336633663366, | |
| "loss": 3.2819, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 12.150791802515137, | |
| "grad_norm": 0.3621805012226105, | |
| "learning_rate": 0.0004544589400116482, | |
| "loss": 3.2831, | |
| "step": 41750 | |
| }, | |
| { | |
| "epoch": 12.165346995808104, | |
| "grad_norm": 0.37465348839759827, | |
| "learning_rate": 0.00045428421665695974, | |
| "loss": 3.2851, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 12.179902189101071, | |
| "grad_norm": 0.3354896903038025, | |
| "learning_rate": 0.00045410949330227133, | |
| "loss": 3.2932, | |
| "step": 41850 | |
| }, | |
| { | |
| "epoch": 12.194457382394038, | |
| "grad_norm": 0.3429812788963318, | |
| "learning_rate": 0.000453934769947583, | |
| "loss": 3.286, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 12.209012575687005, | |
| "grad_norm": 0.36807188391685486, | |
| "learning_rate": 0.0004537600465928946, | |
| "loss": 3.3028, | |
| "step": 41950 | |
| }, | |
| { | |
| "epoch": 12.223567768979972, | |
| "grad_norm": 0.34485989809036255, | |
| "learning_rate": 0.0004535853232382061, | |
| "loss": 3.297, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.223567768979972, | |
| "eval_accuracy": 0.37026154879128453, | |
| "eval_loss": 3.55880069732666, | |
| "eval_runtime": 83.7978, | |
| "eval_samples_per_second": 198.705, | |
| "eval_steps_per_second": 12.423, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 12.23812296227294, | |
| "grad_norm": 0.33578160405158997, | |
| "learning_rate": 0.0004534105998835177, | |
| "loss": 3.2972, | |
| "step": 42050 | |
| }, | |
| { | |
| "epoch": 12.252678155565906, | |
| "grad_norm": 0.3506467640399933, | |
| "learning_rate": 0.0004532358765288293, | |
| "loss": 3.3187, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 12.267233348858873, | |
| "grad_norm": 0.3814696967601776, | |
| "learning_rate": 0.00045306115317414095, | |
| "loss": 3.2947, | |
| "step": 42150 | |
| }, | |
| { | |
| "epoch": 12.28178854215184, | |
| "grad_norm": 0.872353732585907, | |
| "learning_rate": 0.0004528864298194525, | |
| "loss": 3.307, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 12.296343735444808, | |
| "grad_norm": 0.36556026339530945, | |
| "learning_rate": 0.0004527117064647641, | |
| "loss": 3.3018, | |
| "step": 42250 | |
| }, | |
| { | |
| "epoch": 12.310898928737773, | |
| "grad_norm": 0.33834773302078247, | |
| "learning_rate": 0.0004525369831100757, | |
| "loss": 3.322, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 12.32545412203074, | |
| "grad_norm": 0.3654991090297699, | |
| "learning_rate": 0.0004523622597553872, | |
| "loss": 3.3035, | |
| "step": 42350 | |
| }, | |
| { | |
| "epoch": 12.340009315323707, | |
| "grad_norm": 0.3839372396469116, | |
| "learning_rate": 0.0004521875364006988, | |
| "loss": 3.3087, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 12.354564508616674, | |
| "grad_norm": 0.3559876084327698, | |
| "learning_rate": 0.00045201281304601046, | |
| "loss": 3.3175, | |
| "step": 42450 | |
| }, | |
| { | |
| "epoch": 12.369119701909641, | |
| "grad_norm": 0.33983153104782104, | |
| "learning_rate": 0.00045183808969132206, | |
| "loss": 3.3243, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 12.383674895202608, | |
| "grad_norm": 0.343637079000473, | |
| "learning_rate": 0.00045166336633663365, | |
| "loss": 3.3118, | |
| "step": 42550 | |
| }, | |
| { | |
| "epoch": 12.398230088495575, | |
| "grad_norm": 0.3547670841217041, | |
| "learning_rate": 0.0004514886429819452, | |
| "loss": 3.3208, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 12.412785281788542, | |
| "grad_norm": 0.3427077531814575, | |
| "learning_rate": 0.0004513139196272568, | |
| "loss": 3.3174, | |
| "step": 42650 | |
| }, | |
| { | |
| "epoch": 12.42734047508151, | |
| "grad_norm": 0.33483266830444336, | |
| "learning_rate": 0.00045113919627256843, | |
| "loss": 3.3136, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 12.441895668374476, | |
| "grad_norm": 0.34522172808647156, | |
| "learning_rate": 0.00045096447291788003, | |
| "loss": 3.3221, | |
| "step": 42750 | |
| }, | |
| { | |
| "epoch": 12.456450861667443, | |
| "grad_norm": 0.3766433000564575, | |
| "learning_rate": 0.00045078974956319157, | |
| "loss": 3.3242, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 12.47100605496041, | |
| "grad_norm": 0.3491019010543823, | |
| "learning_rate": 0.00045061502620850316, | |
| "loss": 3.324, | |
| "step": 42850 | |
| }, | |
| { | |
| "epoch": 12.485561248253378, | |
| "grad_norm": 0.3642924129962921, | |
| "learning_rate": 0.00045044030285381476, | |
| "loss": 3.3214, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 12.500116441546343, | |
| "grad_norm": 0.3477074205875397, | |
| "learning_rate": 0.0004502655794991263, | |
| "loss": 3.3338, | |
| "step": 42950 | |
| }, | |
| { | |
| "epoch": 12.51467163483931, | |
| "grad_norm": 0.3555394113063812, | |
| "learning_rate": 0.00045009085614443795, | |
| "loss": 3.3204, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.51467163483931, | |
| "eval_accuracy": 0.3707548108921016, | |
| "eval_loss": 3.554856300354004, | |
| "eval_runtime": 83.708, | |
| "eval_samples_per_second": 198.918, | |
| "eval_steps_per_second": 12.436, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 12.529226828132277, | |
| "grad_norm": 0.4010697603225708, | |
| "learning_rate": 0.00044991613278974954, | |
| "loss": 3.3301, | |
| "step": 43050 | |
| }, | |
| { | |
| "epoch": 12.543782021425244, | |
| "grad_norm": 0.3316815495491028, | |
| "learning_rate": 0.00044974140943506113, | |
| "loss": 3.3274, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 12.558337214718211, | |
| "grad_norm": 0.348168283700943, | |
| "learning_rate": 0.0004495666860803727, | |
| "loss": 3.3323, | |
| "step": 43150 | |
| }, | |
| { | |
| "epoch": 12.572892408011178, | |
| "grad_norm": 0.34722900390625, | |
| "learning_rate": 0.00044939196272568427, | |
| "loss": 3.327, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 12.587447601304145, | |
| "grad_norm": 0.36657488346099854, | |
| "learning_rate": 0.00044921723937099586, | |
| "loss": 3.3282, | |
| "step": 43250 | |
| }, | |
| { | |
| "epoch": 12.602002794597112, | |
| "grad_norm": 0.3358417749404907, | |
| "learning_rate": 0.0004490425160163075, | |
| "loss": 3.3307, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 12.61655798789008, | |
| "grad_norm": 0.33212822675704956, | |
| "learning_rate": 0.00044886779266161905, | |
| "loss": 3.3393, | |
| "step": 43350 | |
| }, | |
| { | |
| "epoch": 12.631113181183046, | |
| "grad_norm": 0.3470332622528076, | |
| "learning_rate": 0.00044869306930693065, | |
| "loss": 3.3243, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 12.645668374476013, | |
| "grad_norm": 0.3271242380142212, | |
| "learning_rate": 0.00044851834595224224, | |
| "loss": 3.3375, | |
| "step": 43450 | |
| }, | |
| { | |
| "epoch": 12.66022356776898, | |
| "grad_norm": 0.3358079493045807, | |
| "learning_rate": 0.00044834362259755383, | |
| "loss": 3.3424, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 12.674778761061948, | |
| "grad_norm": 0.34276363253593445, | |
| "learning_rate": 0.00044816889924286543, | |
| "loss": 3.341, | |
| "step": 43550 | |
| }, | |
| { | |
| "epoch": 12.689333954354915, | |
| "grad_norm": 0.3812705874443054, | |
| "learning_rate": 0.000447994175888177, | |
| "loss": 3.336, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 12.703889147647882, | |
| "grad_norm": 0.35825347900390625, | |
| "learning_rate": 0.0004478194525334886, | |
| "loss": 3.3291, | |
| "step": 43650 | |
| }, | |
| { | |
| "epoch": 12.718444340940847, | |
| "grad_norm": 0.3637557029724121, | |
| "learning_rate": 0.0004476447291788002, | |
| "loss": 3.3458, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 12.732999534233814, | |
| "grad_norm": 0.3494272232055664, | |
| "learning_rate": 0.00044747000582411175, | |
| "loss": 3.3386, | |
| "step": 43750 | |
| }, | |
| { | |
| "epoch": 12.747554727526781, | |
| "grad_norm": 0.36245179176330566, | |
| "learning_rate": 0.00044729528246942335, | |
| "loss": 3.3242, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 12.762109920819748, | |
| "grad_norm": 0.34557071328163147, | |
| "learning_rate": 0.000447120559114735, | |
| "loss": 3.328, | |
| "step": 43850 | |
| }, | |
| { | |
| "epoch": 12.776665114112715, | |
| "grad_norm": 0.346245676279068, | |
| "learning_rate": 0.0004469458357600466, | |
| "loss": 3.3513, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 12.791220307405682, | |
| "grad_norm": 0.3590599298477173, | |
| "learning_rate": 0.00044677111240535813, | |
| "loss": 3.3394, | |
| "step": 43950 | |
| }, | |
| { | |
| "epoch": 12.80577550069865, | |
| "grad_norm": 0.3755113184452057, | |
| "learning_rate": 0.0004465963890506697, | |
| "loss": 3.3522, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.80577550069865, | |
| "eval_accuracy": 0.37157515148388215, | |
| "eval_loss": 3.5421037673950195, | |
| "eval_runtime": 83.799, | |
| "eval_samples_per_second": 198.702, | |
| "eval_steps_per_second": 12.423, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 12.820330693991616, | |
| "grad_norm": 0.3785479962825775, | |
| "learning_rate": 0.0004464216656959813, | |
| "loss": 3.3521, | |
| "step": 44050 | |
| }, | |
| { | |
| "epoch": 12.834885887284583, | |
| "grad_norm": 0.32551199197769165, | |
| "learning_rate": 0.00044624694234129297, | |
| "loss": 3.3414, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 12.84944108057755, | |
| "grad_norm": 0.3787110149860382, | |
| "learning_rate": 0.0004460722189866045, | |
| "loss": 3.3425, | |
| "step": 44150 | |
| }, | |
| { | |
| "epoch": 12.863996273870518, | |
| "grad_norm": 0.4027758836746216, | |
| "learning_rate": 0.0004458974956319161, | |
| "loss": 3.3507, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 12.878551467163485, | |
| "grad_norm": 0.33465805649757385, | |
| "learning_rate": 0.0004457227722772277, | |
| "loss": 3.3387, | |
| "step": 44250 | |
| }, | |
| { | |
| "epoch": 12.89310666045645, | |
| "grad_norm": 0.35807907581329346, | |
| "learning_rate": 0.00044554804892253923, | |
| "loss": 3.3413, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 12.907661853749417, | |
| "grad_norm": 0.3467465043067932, | |
| "learning_rate": 0.00044537332556785083, | |
| "loss": 3.3533, | |
| "step": 44350 | |
| }, | |
| { | |
| "epoch": 12.922217047042384, | |
| "grad_norm": 0.3399744927883148, | |
| "learning_rate": 0.0004451986022131625, | |
| "loss": 3.3411, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 12.936772240335351, | |
| "grad_norm": 0.34568867087364197, | |
| "learning_rate": 0.00044502387885847407, | |
| "loss": 3.3501, | |
| "step": 44450 | |
| }, | |
| { | |
| "epoch": 12.951327433628318, | |
| "grad_norm": 0.3443354070186615, | |
| "learning_rate": 0.0004448491555037856, | |
| "loss": 3.3586, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 12.965882626921285, | |
| "grad_norm": 0.3556445837020874, | |
| "learning_rate": 0.0004446744321490972, | |
| "loss": 3.3399, | |
| "step": 44550 | |
| }, | |
| { | |
| "epoch": 12.980437820214252, | |
| "grad_norm": 0.3451250493526459, | |
| "learning_rate": 0.0004444997087944088, | |
| "loss": 3.3492, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 12.99499301350722, | |
| "grad_norm": 0.33421894907951355, | |
| "learning_rate": 0.0004443249854397204, | |
| "loss": 3.341, | |
| "step": 44650 | |
| }, | |
| { | |
| "epoch": 13.009315323707499, | |
| "grad_norm": 0.343733549118042, | |
| "learning_rate": 0.000444150262085032, | |
| "loss": 3.2824, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 13.023870517000466, | |
| "grad_norm": 0.3574601709842682, | |
| "learning_rate": 0.0004439755387303436, | |
| "loss": 3.2413, | |
| "step": 44750 | |
| }, | |
| { | |
| "epoch": 13.038425710293433, | |
| "grad_norm": 0.3487134873867035, | |
| "learning_rate": 0.0004438008153756552, | |
| "loss": 3.2501, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 13.0529809035864, | |
| "grad_norm": 0.38518014550209045, | |
| "learning_rate": 0.00044362609202096677, | |
| "loss": 3.2581, | |
| "step": 44850 | |
| }, | |
| { | |
| "epoch": 13.067536096879367, | |
| "grad_norm": 0.3497874140739441, | |
| "learning_rate": 0.0004434513686662783, | |
| "loss": 3.2545, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 13.082091290172334, | |
| "grad_norm": 0.36562028527259827, | |
| "learning_rate": 0.00044327664531158996, | |
| "loss": 3.2447, | |
| "step": 44950 | |
| }, | |
| { | |
| "epoch": 13.0966464834653, | |
| "grad_norm": 0.3318837881088257, | |
| "learning_rate": 0.00044310192195690155, | |
| "loss": 3.2624, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.0966464834653, | |
| "eval_accuracy": 0.3708592926666135, | |
| "eval_loss": 3.5579721927642822, | |
| "eval_runtime": 83.7497, | |
| "eval_samples_per_second": 198.819, | |
| "eval_steps_per_second": 12.43, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 13.111201676758267, | |
| "grad_norm": 0.3484169542789459, | |
| "learning_rate": 0.00044292719860221315, | |
| "loss": 3.2619, | |
| "step": 45050 | |
| }, | |
| { | |
| "epoch": 13.125756870051234, | |
| "grad_norm": 0.35762280225753784, | |
| "learning_rate": 0.0004427524752475247, | |
| "loss": 3.276, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 13.1403120633442, | |
| "grad_norm": 0.3637618124485016, | |
| "learning_rate": 0.0004425777518928363, | |
| "loss": 3.2701, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 13.154867256637168, | |
| "grad_norm": 0.38000133633613586, | |
| "learning_rate": 0.0004424030285381479, | |
| "loss": 3.2683, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 13.169422449930135, | |
| "grad_norm": 0.3983210027217865, | |
| "learning_rate": 0.0004422283051834595, | |
| "loss": 3.2642, | |
| "step": 45250 | |
| }, | |
| { | |
| "epoch": 13.183977643223102, | |
| "grad_norm": 0.3654680848121643, | |
| "learning_rate": 0.00044205358182877107, | |
| "loss": 3.2794, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 13.198532836516069, | |
| "grad_norm": 0.37247559428215027, | |
| "learning_rate": 0.00044187885847408266, | |
| "loss": 3.2716, | |
| "step": 45350 | |
| }, | |
| { | |
| "epoch": 13.213088029809036, | |
| "grad_norm": 0.38887423276901245, | |
| "learning_rate": 0.00044170413511939425, | |
| "loss": 3.2879, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 13.227643223102003, | |
| "grad_norm": 0.3559700846672058, | |
| "learning_rate": 0.0004415294117647058, | |
| "loss": 3.2808, | |
| "step": 45450 | |
| }, | |
| { | |
| "epoch": 13.24219841639497, | |
| "grad_norm": 0.36184707283973694, | |
| "learning_rate": 0.00044135468841001744, | |
| "loss": 3.2802, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 13.256753609687937, | |
| "grad_norm": 0.3328874707221985, | |
| "learning_rate": 0.00044117996505532904, | |
| "loss": 3.2791, | |
| "step": 45550 | |
| }, | |
| { | |
| "epoch": 13.271308802980904, | |
| "grad_norm": 0.36260300874710083, | |
| "learning_rate": 0.00044100524170064063, | |
| "loss": 3.2872, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 13.285863996273871, | |
| "grad_norm": 0.34145623445510864, | |
| "learning_rate": 0.0004408305183459522, | |
| "loss": 3.2747, | |
| "step": 45650 | |
| }, | |
| { | |
| "epoch": 13.300419189566837, | |
| "grad_norm": 0.34585994482040405, | |
| "learning_rate": 0.00044065579499126377, | |
| "loss": 3.2867, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 13.314974382859804, | |
| "grad_norm": 0.35756462812423706, | |
| "learning_rate": 0.00044048107163657536, | |
| "loss": 3.3042, | |
| "step": 45750 | |
| }, | |
| { | |
| "epoch": 13.32952957615277, | |
| "grad_norm": 0.3436504602432251, | |
| "learning_rate": 0.000440306348281887, | |
| "loss": 3.2851, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 13.344084769445738, | |
| "grad_norm": 0.3949281871318817, | |
| "learning_rate": 0.0004401316249271986, | |
| "loss": 3.2954, | |
| "step": 45850 | |
| }, | |
| { | |
| "epoch": 13.358639962738705, | |
| "grad_norm": 0.336416631937027, | |
| "learning_rate": 0.00043995690157251014, | |
| "loss": 3.292, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 13.373195156031672, | |
| "grad_norm": 0.3523288071155548, | |
| "learning_rate": 0.00043978217821782174, | |
| "loss": 3.289, | |
| "step": 45950 | |
| }, | |
| { | |
| "epoch": 13.387750349324639, | |
| "grad_norm": 0.37476205825805664, | |
| "learning_rate": 0.00043960745486313333, | |
| "loss": 3.295, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.387750349324639, | |
| "eval_accuracy": 0.3713348081443132, | |
| "eval_loss": 3.5528359413146973, | |
| "eval_runtime": 83.7236, | |
| "eval_samples_per_second": 198.881, | |
| "eval_steps_per_second": 12.434, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 13.402305542617606, | |
| "grad_norm": 0.3564622104167938, | |
| "learning_rate": 0.00043943273150844487, | |
| "loss": 3.3042, | |
| "step": 46050 | |
| }, | |
| { | |
| "epoch": 13.416860735910573, | |
| "grad_norm": 0.35462141036987305, | |
| "learning_rate": 0.0004392580081537565, | |
| "loss": 3.3046, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 13.43141592920354, | |
| "grad_norm": 0.34062400460243225, | |
| "learning_rate": 0.0004390832847990681, | |
| "loss": 3.302, | |
| "step": 46150 | |
| }, | |
| { | |
| "epoch": 13.445971122496507, | |
| "grad_norm": 0.3310416638851166, | |
| "learning_rate": 0.0004389085614443797, | |
| "loss": 3.2968, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 13.460526315789474, | |
| "grad_norm": 0.37062373757362366, | |
| "learning_rate": 0.00043873383808969125, | |
| "loss": 3.3017, | |
| "step": 46250 | |
| }, | |
| { | |
| "epoch": 13.475081509082441, | |
| "grad_norm": 0.3724871277809143, | |
| "learning_rate": 0.00043855911473500284, | |
| "loss": 3.301, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 13.489636702375407, | |
| "grad_norm": 0.35389065742492676, | |
| "learning_rate": 0.0004383843913803145, | |
| "loss": 3.3257, | |
| "step": 46350 | |
| }, | |
| { | |
| "epoch": 13.504191895668374, | |
| "grad_norm": 0.3547104299068451, | |
| "learning_rate": 0.0004382096680256261, | |
| "loss": 3.3167, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 13.51874708896134, | |
| "grad_norm": 0.3604714274406433, | |
| "learning_rate": 0.0004380349446709376, | |
| "loss": 3.3083, | |
| "step": 46450 | |
| }, | |
| { | |
| "epoch": 13.533302282254308, | |
| "grad_norm": 0.37563356757164, | |
| "learning_rate": 0.0004378602213162492, | |
| "loss": 3.3068, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 13.547857475547275, | |
| "grad_norm": 0.3432236611843109, | |
| "learning_rate": 0.0004376854979615608, | |
| "loss": 3.3003, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 13.562412668840242, | |
| "grad_norm": 0.3589812219142914, | |
| "learning_rate": 0.0004375107746068724, | |
| "loss": 3.3117, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 13.576967862133209, | |
| "grad_norm": 0.3608124554157257, | |
| "learning_rate": 0.000437336051252184, | |
| "loss": 3.3125, | |
| "step": 46650 | |
| }, | |
| { | |
| "epoch": 13.591523055426176, | |
| "grad_norm": 0.3663826584815979, | |
| "learning_rate": 0.0004371613278974956, | |
| "loss": 3.3164, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 13.606078248719143, | |
| "grad_norm": 0.34340113401412964, | |
| "learning_rate": 0.0004369866045428072, | |
| "loss": 3.3019, | |
| "step": 46750 | |
| }, | |
| { | |
| "epoch": 13.62063344201211, | |
| "grad_norm": 0.3766665458679199, | |
| "learning_rate": 0.0004368118811881188, | |
| "loss": 3.3173, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 13.635188635305077, | |
| "grad_norm": 0.33419665694236755, | |
| "learning_rate": 0.0004366371578334303, | |
| "loss": 3.3167, | |
| "step": 46850 | |
| }, | |
| { | |
| "epoch": 13.649743828598044, | |
| "grad_norm": 0.353169322013855, | |
| "learning_rate": 0.000436462434478742, | |
| "loss": 3.3181, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 13.664299021891011, | |
| "grad_norm": 0.33942270278930664, | |
| "learning_rate": 0.00043628771112405357, | |
| "loss": 3.3259, | |
| "step": 46950 | |
| }, | |
| { | |
| "epoch": 13.678854215183978, | |
| "grad_norm": 0.33569732308387756, | |
| "learning_rate": 0.00043611298776936516, | |
| "loss": 3.3167, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.678854215183978, | |
| "eval_accuracy": 0.37179210688967396, | |
| "eval_loss": 3.5467529296875, | |
| "eval_runtime": 83.7185, | |
| "eval_samples_per_second": 198.893, | |
| "eval_steps_per_second": 12.435, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 13.693409408476944, | |
| "grad_norm": 0.36921167373657227, | |
| "learning_rate": 0.0004359382644146767, | |
| "loss": 3.3153, | |
| "step": 47050 | |
| }, | |
| { | |
| "epoch": 13.70796460176991, | |
| "grad_norm": 0.36710643768310547, | |
| "learning_rate": 0.0004357635410599883, | |
| "loss": 3.3332, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 13.722519795062878, | |
| "grad_norm": 0.3771568238735199, | |
| "learning_rate": 0.0004355888177052999, | |
| "loss": 3.3318, | |
| "step": 47150 | |
| }, | |
| { | |
| "epoch": 13.737074988355845, | |
| "grad_norm": 0.34348955750465393, | |
| "learning_rate": 0.00043541409435061154, | |
| "loss": 3.3244, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 13.751630181648812, | |
| "grad_norm": 0.359382301568985, | |
| "learning_rate": 0.0004352393709959231, | |
| "loss": 3.3151, | |
| "step": 47250 | |
| }, | |
| { | |
| "epoch": 13.766185374941779, | |
| "grad_norm": 0.3418557047843933, | |
| "learning_rate": 0.0004350646476412347, | |
| "loss": 3.329, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 13.780740568234746, | |
| "grad_norm": 0.3581181466579437, | |
| "learning_rate": 0.00043488992428654627, | |
| "loss": 3.3363, | |
| "step": 47350 | |
| }, | |
| { | |
| "epoch": 13.795295761527713, | |
| "grad_norm": 0.3581472337245941, | |
| "learning_rate": 0.0004347152009318578, | |
| "loss": 3.3256, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 13.80985095482068, | |
| "grad_norm": 0.34938347339630127, | |
| "learning_rate": 0.00043454047757716946, | |
| "loss": 3.3328, | |
| "step": 47450 | |
| }, | |
| { | |
| "epoch": 13.824406148113647, | |
| "grad_norm": 0.3522000312805176, | |
| "learning_rate": 0.00043436575422248105, | |
| "loss": 3.3267, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 13.838961341406614, | |
| "grad_norm": 0.39137163758277893, | |
| "learning_rate": 0.00043419103086779265, | |
| "loss": 3.3342, | |
| "step": 47550 | |
| }, | |
| { | |
| "epoch": 13.853516534699581, | |
| "grad_norm": 0.32812169194221497, | |
| "learning_rate": 0.0004340163075131042, | |
| "loss": 3.3205, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 13.868071727992549, | |
| "grad_norm": 0.36968374252319336, | |
| "learning_rate": 0.0004338415841584158, | |
| "loss": 3.3415, | |
| "step": 47650 | |
| }, | |
| { | |
| "epoch": 13.882626921285514, | |
| "grad_norm": 0.3799431025981903, | |
| "learning_rate": 0.0004336668608037274, | |
| "loss": 3.3233, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 13.89718211457848, | |
| "grad_norm": 0.36741307377815247, | |
| "learning_rate": 0.000433492137449039, | |
| "loss": 3.3306, | |
| "step": 47750 | |
| }, | |
| { | |
| "epoch": 13.911737307871448, | |
| "grad_norm": 0.36500704288482666, | |
| "learning_rate": 0.00043331741409435056, | |
| "loss": 3.331, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 13.926292501164415, | |
| "grad_norm": 0.3221661448478699, | |
| "learning_rate": 0.00043314269073966216, | |
| "loss": 3.3341, | |
| "step": 47850 | |
| }, | |
| { | |
| "epoch": 13.940847694457382, | |
| "grad_norm": 0.386502206325531, | |
| "learning_rate": 0.00043296796738497375, | |
| "loss": 3.3205, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 13.955402887750349, | |
| "grad_norm": 0.36427491903305054, | |
| "learning_rate": 0.00043279324403028535, | |
| "loss": 3.348, | |
| "step": 47950 | |
| }, | |
| { | |
| "epoch": 13.969958081043316, | |
| "grad_norm": 0.36302849650382996, | |
| "learning_rate": 0.0004326185206755969, | |
| "loss": 3.3347, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.969958081043316, | |
| "eval_accuracy": 0.3724107706253663, | |
| "eval_loss": 3.5372579097747803, | |
| "eval_runtime": 83.7363, | |
| "eval_samples_per_second": 198.85, | |
| "eval_steps_per_second": 12.432, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 13.984513274336283, | |
| "grad_norm": 0.369755357503891, | |
| "learning_rate": 0.00043244379732090854, | |
| "loss": 3.3425, | |
| "step": 48050 | |
| }, | |
| { | |
| "epoch": 13.99906846762925, | |
| "grad_norm": 0.3400190472602844, | |
| "learning_rate": 0.00043226907396622013, | |
| "loss": 3.3286, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 14.01339077782953, | |
| "grad_norm": 0.36831361055374146, | |
| "learning_rate": 0.0004320943506115317, | |
| "loss": 3.2305, | |
| "step": 48150 | |
| }, | |
| { | |
| "epoch": 14.027945971122497, | |
| "grad_norm": 0.3354513645172119, | |
| "learning_rate": 0.00043191962725684326, | |
| "loss": 3.2128, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 14.042501164415464, | |
| "grad_norm": 0.37795165181159973, | |
| "learning_rate": 0.00043174490390215486, | |
| "loss": 3.2406, | |
| "step": 48250 | |
| }, | |
| { | |
| "epoch": 14.057056357708431, | |
| "grad_norm": 0.3384867310523987, | |
| "learning_rate": 0.0004315701805474665, | |
| "loss": 3.2299, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 14.071611551001398, | |
| "grad_norm": 0.3573961853981018, | |
| "learning_rate": 0.0004313954571927781, | |
| "loss": 3.2339, | |
| "step": 48350 | |
| }, | |
| { | |
| "epoch": 14.086166744294363, | |
| "grad_norm": 0.34502115845680237, | |
| "learning_rate": 0.00043122073383808964, | |
| "loss": 3.2476, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 14.10072193758733, | |
| "grad_norm": 0.3816913664340973, | |
| "learning_rate": 0.00043104601048340124, | |
| "loss": 3.2463, | |
| "step": 48450 | |
| }, | |
| { | |
| "epoch": 14.115277130880298, | |
| "grad_norm": 0.33535298705101013, | |
| "learning_rate": 0.00043087128712871283, | |
| "loss": 3.2523, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 14.129832324173265, | |
| "grad_norm": 0.34241580963134766, | |
| "learning_rate": 0.00043069656377402437, | |
| "loss": 3.2554, | |
| "step": 48550 | |
| }, | |
| { | |
| "epoch": 14.144387517466232, | |
| "grad_norm": 0.3590269982814789, | |
| "learning_rate": 0.000430521840419336, | |
| "loss": 3.2468, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 14.158942710759199, | |
| "grad_norm": 0.3646739423274994, | |
| "learning_rate": 0.0004303471170646476, | |
| "loss": 3.2522, | |
| "step": 48650 | |
| }, | |
| { | |
| "epoch": 14.173497904052166, | |
| "grad_norm": 0.34746336936950684, | |
| "learning_rate": 0.0004301723937099592, | |
| "loss": 3.2552, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 14.188053097345133, | |
| "grad_norm": 0.3597225844860077, | |
| "learning_rate": 0.00042999767035527075, | |
| "loss": 3.2673, | |
| "step": 48750 | |
| }, | |
| { | |
| "epoch": 14.2026082906381, | |
| "grad_norm": 0.37786567211151123, | |
| "learning_rate": 0.00042982294700058234, | |
| "loss": 3.2619, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 14.217163483931067, | |
| "grad_norm": 0.38890740275382996, | |
| "learning_rate": 0.000429648223645894, | |
| "loss": 3.2678, | |
| "step": 48850 | |
| }, | |
| { | |
| "epoch": 14.231718677224034, | |
| "grad_norm": 0.337555855512619, | |
| "learning_rate": 0.0004294735002912056, | |
| "loss": 3.2737, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 14.246273870517001, | |
| "grad_norm": 0.40954357385635376, | |
| "learning_rate": 0.0004292987769365172, | |
| "loss": 3.2636, | |
| "step": 48950 | |
| }, | |
| { | |
| "epoch": 14.260829063809968, | |
| "grad_norm": 0.37603989243507385, | |
| "learning_rate": 0.0004291240535818287, | |
| "loss": 3.2663, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.260829063809968, | |
| "eval_accuracy": 0.3716564803792277, | |
| "eval_loss": 3.5511670112609863, | |
| "eval_runtime": 83.7051, | |
| "eval_samples_per_second": 198.925, | |
| "eval_steps_per_second": 12.437, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 14.275384257102935, | |
| "grad_norm": 0.3539530038833618, | |
| "learning_rate": 0.0004289493302271403, | |
| "loss": 3.2806, | |
| "step": 49050 | |
| }, | |
| { | |
| "epoch": 14.2899394503959, | |
| "grad_norm": 0.36623480916023254, | |
| "learning_rate": 0.0004287746068724519, | |
| "loss": 3.2841, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 14.304494643688868, | |
| "grad_norm": 0.34146666526794434, | |
| "learning_rate": 0.00042859988351776356, | |
| "loss": 3.2669, | |
| "step": 49150 | |
| }, | |
| { | |
| "epoch": 14.319049836981835, | |
| "grad_norm": 0.40605178475379944, | |
| "learning_rate": 0.0004284251601630751, | |
| "loss": 3.2837, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 14.333605030274802, | |
| "grad_norm": 0.3694278299808502, | |
| "learning_rate": 0.0004282504368083867, | |
| "loss": 3.2685, | |
| "step": 49250 | |
| }, | |
| { | |
| "epoch": 14.348160223567769, | |
| "grad_norm": 0.3507702052593231, | |
| "learning_rate": 0.0004280757134536983, | |
| "loss": 3.2864, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 14.362715416860736, | |
| "grad_norm": 0.37406304478645325, | |
| "learning_rate": 0.0004279009900990098, | |
| "loss": 3.2824, | |
| "step": 49350 | |
| }, | |
| { | |
| "epoch": 14.377270610153703, | |
| "grad_norm": 0.35801663994789124, | |
| "learning_rate": 0.0004277262667443214, | |
| "loss": 3.2816, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 14.39182580344667, | |
| "grad_norm": 0.332137793302536, | |
| "learning_rate": 0.00042755154338963307, | |
| "loss": 3.2889, | |
| "step": 49450 | |
| }, | |
| { | |
| "epoch": 14.406380996739637, | |
| "grad_norm": 0.361145555973053, | |
| "learning_rate": 0.00042737682003494466, | |
| "loss": 3.2947, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 14.420936190032604, | |
| "grad_norm": 0.37780454754829407, | |
| "learning_rate": 0.0004272020966802562, | |
| "loss": 3.2983, | |
| "step": 49550 | |
| }, | |
| { | |
| "epoch": 14.435491383325571, | |
| "grad_norm": 0.34408506751060486, | |
| "learning_rate": 0.0004270273733255678, | |
| "loss": 3.2849, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 14.450046576618538, | |
| "grad_norm": 0.3806084096431732, | |
| "learning_rate": 0.0004268526499708794, | |
| "loss": 3.2885, | |
| "step": 49650 | |
| }, | |
| { | |
| "epoch": 14.464601769911505, | |
| "grad_norm": 0.3702000081539154, | |
| "learning_rate": 0.00042667792661619104, | |
| "loss": 3.2921, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 14.47915696320447, | |
| "grad_norm": 0.3677707612514496, | |
| "learning_rate": 0.0004265032032615026, | |
| "loss": 3.2925, | |
| "step": 49750 | |
| }, | |
| { | |
| "epoch": 14.493712156497438, | |
| "grad_norm": 0.34538209438323975, | |
| "learning_rate": 0.0004263284799068142, | |
| "loss": 3.2906, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 14.508267349790405, | |
| "grad_norm": 0.3703807294368744, | |
| "learning_rate": 0.00042615375655212577, | |
| "loss": 3.2953, | |
| "step": 49850 | |
| }, | |
| { | |
| "epoch": 14.522822543083372, | |
| "grad_norm": 0.36870095133781433, | |
| "learning_rate": 0.00042597903319743736, | |
| "loss": 3.2837, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 14.537377736376339, | |
| "grad_norm": 0.3427352011203766, | |
| "learning_rate": 0.0004258043098427489, | |
| "loss": 3.2889, | |
| "step": 49950 | |
| }, | |
| { | |
| "epoch": 14.551932929669306, | |
| "grad_norm": 0.3615899980068207, | |
| "learning_rate": 0.00042562958648806055, | |
| "loss": 3.2948, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.551932929669306, | |
| "eval_accuracy": 0.37187907709567936, | |
| "eval_loss": 3.544443130493164, | |
| "eval_runtime": 83.7261, | |
| "eval_samples_per_second": 198.875, | |
| "eval_steps_per_second": 12.433, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 14.566488122962273, | |
| "grad_norm": 0.37290671467781067, | |
| "learning_rate": 0.00042545486313337214, | |
| "loss": 3.2855, | |
| "step": 50050 | |
| }, | |
| { | |
| "epoch": 14.58104331625524, | |
| "grad_norm": 0.3681708872318268, | |
| "learning_rate": 0.00042528013977868374, | |
| "loss": 3.2974, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 14.595598509548207, | |
| "grad_norm": 0.3454001247882843, | |
| "learning_rate": 0.0004251054164239953, | |
| "loss": 3.289, | |
| "step": 50150 | |
| }, | |
| { | |
| "epoch": 14.610153702841174, | |
| "grad_norm": 0.4051375091075897, | |
| "learning_rate": 0.0004249306930693069, | |
| "loss": 3.3147, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 14.624708896134141, | |
| "grad_norm": 0.3999985158443451, | |
| "learning_rate": 0.0004247559697146185, | |
| "loss": 3.3001, | |
| "step": 50250 | |
| }, | |
| { | |
| "epoch": 14.639264089427108, | |
| "grad_norm": 0.3676909804344177, | |
| "learning_rate": 0.0004245812463599301, | |
| "loss": 3.2985, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 14.653819282720075, | |
| "grad_norm": 0.382487952709198, | |
| "learning_rate": 0.00042440652300524166, | |
| "loss": 3.2883, | |
| "step": 50350 | |
| }, | |
| { | |
| "epoch": 14.668374476013042, | |
| "grad_norm": 0.3541913330554962, | |
| "learning_rate": 0.00042423179965055325, | |
| "loss": 3.3014, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 14.682929669306008, | |
| "grad_norm": 0.35107702016830444, | |
| "learning_rate": 0.00042405707629586484, | |
| "loss": 3.3062, | |
| "step": 50450 | |
| }, | |
| { | |
| "epoch": 14.697484862598975, | |
| "grad_norm": 0.3456815183162689, | |
| "learning_rate": 0.0004238823529411764, | |
| "loss": 3.3113, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 14.712040055891942, | |
| "grad_norm": 0.3480520248413086, | |
| "learning_rate": 0.00042370762958648803, | |
| "loss": 3.3049, | |
| "step": 50550 | |
| }, | |
| { | |
| "epoch": 14.726595249184909, | |
| "grad_norm": 0.35998451709747314, | |
| "learning_rate": 0.00042353290623179963, | |
| "loss": 3.304, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 14.741150442477876, | |
| "grad_norm": 0.3671087920665741, | |
| "learning_rate": 0.0004233581828771112, | |
| "loss": 3.3035, | |
| "step": 50650 | |
| }, | |
| { | |
| "epoch": 14.755705635770843, | |
| "grad_norm": 0.34634125232696533, | |
| "learning_rate": 0.00042318345952242276, | |
| "loss": 3.3165, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 14.77026082906381, | |
| "grad_norm": 0.36663252115249634, | |
| "learning_rate": 0.00042300873616773436, | |
| "loss": 3.3099, | |
| "step": 50750 | |
| }, | |
| { | |
| "epoch": 14.784816022356777, | |
| "grad_norm": 0.38535112142562866, | |
| "learning_rate": 0.00042283401281304595, | |
| "loss": 3.3102, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 14.799371215649744, | |
| "grad_norm": 0.33679986000061035, | |
| "learning_rate": 0.0004226592894583576, | |
| "loss": 3.3096, | |
| "step": 50850 | |
| }, | |
| { | |
| "epoch": 14.813926408942711, | |
| "grad_norm": 0.38355258107185364, | |
| "learning_rate": 0.00042248456610366914, | |
| "loss": 3.3042, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 14.828481602235678, | |
| "grad_norm": 0.36840131878852844, | |
| "learning_rate": 0.00042230984274898073, | |
| "loss": 3.3181, | |
| "step": 50950 | |
| }, | |
| { | |
| "epoch": 14.843036795528645, | |
| "grad_norm": 0.34980928897857666, | |
| "learning_rate": 0.00042213511939429233, | |
| "loss": 3.3313, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.843036795528645, | |
| "eval_accuracy": 0.37271058278147406, | |
| "eval_loss": 3.535724639892578, | |
| "eval_runtime": 83.8125, | |
| "eval_samples_per_second": 198.67, | |
| "eval_steps_per_second": 12.421, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 14.857591988821612, | |
| "grad_norm": 0.35362154245376587, | |
| "learning_rate": 0.0004219603960396039, | |
| "loss": 3.3118, | |
| "step": 51050 | |
| }, | |
| { | |
| "epoch": 14.872147182114578, | |
| "grad_norm": 0.35500141978263855, | |
| "learning_rate": 0.0004217856726849155, | |
| "loss": 3.3195, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 14.886702375407545, | |
| "grad_norm": 0.3626366853713989, | |
| "learning_rate": 0.0004216109493302271, | |
| "loss": 3.3238, | |
| "step": 51150 | |
| }, | |
| { | |
| "epoch": 14.901257568700512, | |
| "grad_norm": 0.37499237060546875, | |
| "learning_rate": 0.0004214362259755387, | |
| "loss": 3.312, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 14.915812761993479, | |
| "grad_norm": 0.37007656693458557, | |
| "learning_rate": 0.0004212615026208503, | |
| "loss": 3.3205, | |
| "step": 51250 | |
| }, | |
| { | |
| "epoch": 14.930367955286446, | |
| "grad_norm": 0.3392346203327179, | |
| "learning_rate": 0.00042108677926616184, | |
| "loss": 3.3214, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 14.944923148579413, | |
| "grad_norm": 0.3695128262042999, | |
| "learning_rate": 0.00042091205591147343, | |
| "loss": 3.3206, | |
| "step": 51350 | |
| }, | |
| { | |
| "epoch": 14.95947834187238, | |
| "grad_norm": 0.41312822699546814, | |
| "learning_rate": 0.0004207373325567851, | |
| "loss": 3.3258, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 14.974033535165347, | |
| "grad_norm": 0.3796527683734894, | |
| "learning_rate": 0.0004205626092020967, | |
| "loss": 3.3156, | |
| "step": 51450 | |
| }, | |
| { | |
| "epoch": 14.988588728458314, | |
| "grad_norm": 0.3707231879234314, | |
| "learning_rate": 0.0004203878858474082, | |
| "loss": 3.3018, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 15.002911038658594, | |
| "grad_norm": 0.37344539165496826, | |
| "learning_rate": 0.0004202131624927198, | |
| "loss": 3.3081, | |
| "step": 51550 | |
| }, | |
| { | |
| "epoch": 15.01746623195156, | |
| "grad_norm": 0.36107468605041504, | |
| "learning_rate": 0.0004200384391380314, | |
| "loss": 3.2218, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 15.032021425244528, | |
| "grad_norm": 0.3806946277618408, | |
| "learning_rate": 0.00041986371578334305, | |
| "loss": 3.2066, | |
| "step": 51650 | |
| }, | |
| { | |
| "epoch": 15.046576618537495, | |
| "grad_norm": 0.3937886357307434, | |
| "learning_rate": 0.0004196889924286546, | |
| "loss": 3.2155, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 15.06113181183046, | |
| "grad_norm": 0.35221195220947266, | |
| "learning_rate": 0.0004195142690739662, | |
| "loss": 3.2301, | |
| "step": 51750 | |
| }, | |
| { | |
| "epoch": 15.075687005123427, | |
| "grad_norm": 0.3820745348930359, | |
| "learning_rate": 0.0004193395457192778, | |
| "loss": 3.223, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 15.090242198416394, | |
| "grad_norm": 0.38950881361961365, | |
| "learning_rate": 0.0004191648223645893, | |
| "loss": 3.2315, | |
| "step": 51850 | |
| }, | |
| { | |
| "epoch": 15.104797391709361, | |
| "grad_norm": 0.361758291721344, | |
| "learning_rate": 0.0004189900990099009, | |
| "loss": 3.2323, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 15.119352585002328, | |
| "grad_norm": 0.37369921803474426, | |
| "learning_rate": 0.00041881537565521256, | |
| "loss": 3.2329, | |
| "step": 51950 | |
| }, | |
| { | |
| "epoch": 15.133907778295296, | |
| "grad_norm": 0.3731876611709595, | |
| "learning_rate": 0.00041864065230052416, | |
| "loss": 3.2304, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.133907778295296, | |
| "eval_accuracy": 0.37196686999282264, | |
| "eval_loss": 3.5524168014526367, | |
| "eval_runtime": 83.0314, | |
| "eval_samples_per_second": 200.539, | |
| "eval_steps_per_second": 12.537, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 15.148462971588263, | |
| "grad_norm": 0.3429223597049713, | |
| "learning_rate": 0.00041846592894583575, | |
| "loss": 3.2405, | |
| "step": 52050 | |
| }, | |
| { | |
| "epoch": 15.16301816488123, | |
| "grad_norm": 0.3717568814754486, | |
| "learning_rate": 0.0004182912055911473, | |
| "loss": 3.2387, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 15.177573358174197, | |
| "grad_norm": 0.36367765069007874, | |
| "learning_rate": 0.0004181164822364589, | |
| "loss": 3.2321, | |
| "step": 52150 | |
| }, | |
| { | |
| "epoch": 15.192128551467164, | |
| "grad_norm": 0.3607712984085083, | |
| "learning_rate": 0.0004179417588817705, | |
| "loss": 3.2474, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 15.20668374476013, | |
| "grad_norm": 0.35510334372520447, | |
| "learning_rate": 0.00041776703552708213, | |
| "loss": 3.2515, | |
| "step": 52250 | |
| }, | |
| { | |
| "epoch": 15.221238938053098, | |
| "grad_norm": 0.37096700072288513, | |
| "learning_rate": 0.00041759231217239367, | |
| "loss": 3.2471, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 15.235794131346065, | |
| "grad_norm": 0.37920671701431274, | |
| "learning_rate": 0.00041741758881770527, | |
| "loss": 3.25, | |
| "step": 52350 | |
| }, | |
| { | |
| "epoch": 15.250349324639032, | |
| "grad_norm": 0.3903672397136688, | |
| "learning_rate": 0.00041724286546301686, | |
| "loss": 3.2646, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 15.264904517931997, | |
| "grad_norm": 0.37260299921035767, | |
| "learning_rate": 0.0004170681421083284, | |
| "loss": 3.2622, | |
| "step": 52450 | |
| }, | |
| { | |
| "epoch": 15.279459711224964, | |
| "grad_norm": 0.3678816556930542, | |
| "learning_rate": 0.00041689341875364005, | |
| "loss": 3.2563, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 15.294014904517931, | |
| "grad_norm": 0.37712210416793823, | |
| "learning_rate": 0.00041671869539895164, | |
| "loss": 3.2467, | |
| "step": 52550 | |
| }, | |
| { | |
| "epoch": 15.308570097810899, | |
| "grad_norm": 0.35957497358322144, | |
| "learning_rate": 0.00041654397204426324, | |
| "loss": 3.2664, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 15.323125291103866, | |
| "grad_norm": 0.377457857131958, | |
| "learning_rate": 0.0004163692486895748, | |
| "loss": 3.2701, | |
| "step": 52650 | |
| }, | |
| { | |
| "epoch": 15.337680484396833, | |
| "grad_norm": 0.37984439730644226, | |
| "learning_rate": 0.00041619452533488637, | |
| "loss": 3.272, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 15.3522356776898, | |
| "grad_norm": 0.35283443331718445, | |
| "learning_rate": 0.00041601980198019797, | |
| "loss": 3.2488, | |
| "step": 52750 | |
| }, | |
| { | |
| "epoch": 15.366790870982767, | |
| "grad_norm": 0.37909993529319763, | |
| "learning_rate": 0.0004158450786255096, | |
| "loss": 3.2603, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 15.381346064275734, | |
| "grad_norm": 0.3768189549446106, | |
| "learning_rate": 0.00041567035527082115, | |
| "loss": 3.2657, | |
| "step": 52850 | |
| }, | |
| { | |
| "epoch": 15.3959012575687, | |
| "grad_norm": 0.3707028329372406, | |
| "learning_rate": 0.00041549563191613275, | |
| "loss": 3.2668, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 15.410456450861668, | |
| "grad_norm": 0.3549076020717621, | |
| "learning_rate": 0.00041532090856144434, | |
| "loss": 3.2655, | |
| "step": 52950 | |
| }, | |
| { | |
| "epoch": 15.425011644154635, | |
| "grad_norm": 0.39305442571640015, | |
| "learning_rate": 0.00041514618520675594, | |
| "loss": 3.2744, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.425011644154635, | |
| "eval_accuracy": 0.3723168663083416, | |
| "eval_loss": 3.546820640563965, | |
| "eval_runtime": 83.1139, | |
| "eval_samples_per_second": 200.34, | |
| "eval_steps_per_second": 12.525, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 15.439566837447602, | |
| "grad_norm": 0.3791975677013397, | |
| "learning_rate": 0.00041497146185206753, | |
| "loss": 3.286, | |
| "step": 53050 | |
| }, | |
| { | |
| "epoch": 15.454122030740567, | |
| "grad_norm": 0.3472748100757599, | |
| "learning_rate": 0.0004147967384973791, | |
| "loss": 3.2566, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 15.468677224033534, | |
| "grad_norm": 0.38916853070259094, | |
| "learning_rate": 0.0004146220151426907, | |
| "loss": 3.2736, | |
| "step": 53150 | |
| }, | |
| { | |
| "epoch": 15.483232417326501, | |
| "grad_norm": 0.3874084949493408, | |
| "learning_rate": 0.0004144472917880023, | |
| "loss": 3.2863, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 15.497787610619469, | |
| "grad_norm": 0.3649914860725403, | |
| "learning_rate": 0.00041427256843331385, | |
| "loss": 3.2678, | |
| "step": 53250 | |
| }, | |
| { | |
| "epoch": 15.512342803912436, | |
| "grad_norm": 0.38252490758895874, | |
| "learning_rate": 0.00041409784507862545, | |
| "loss": 3.2846, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 15.526897997205403, | |
| "grad_norm": 0.3822208046913147, | |
| "learning_rate": 0.0004139231217239371, | |
| "loss": 3.2732, | |
| "step": 53350 | |
| }, | |
| { | |
| "epoch": 15.54145319049837, | |
| "grad_norm": 0.35340163111686707, | |
| "learning_rate": 0.0004137483983692487, | |
| "loss": 3.2774, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 15.556008383791337, | |
| "grad_norm": 0.3682171702384949, | |
| "learning_rate": 0.00041357367501456023, | |
| "loss": 3.2933, | |
| "step": 53450 | |
| }, | |
| { | |
| "epoch": 15.570563577084304, | |
| "grad_norm": 0.37386003136634827, | |
| "learning_rate": 0.0004133989516598718, | |
| "loss": 3.2866, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 15.585118770377271, | |
| "grad_norm": 0.34851980209350586, | |
| "learning_rate": 0.0004132242283051834, | |
| "loss": 3.2759, | |
| "step": 53550 | |
| }, | |
| { | |
| "epoch": 15.599673963670238, | |
| "grad_norm": 0.3770555853843689, | |
| "learning_rate": 0.00041304950495049496, | |
| "loss": 3.2859, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 15.614229156963205, | |
| "grad_norm": 0.37687060236930847, | |
| "learning_rate": 0.0004128747815958066, | |
| "loss": 3.2854, | |
| "step": 53650 | |
| }, | |
| { | |
| "epoch": 15.628784350256172, | |
| "grad_norm": 0.38373517990112305, | |
| "learning_rate": 0.0004127000582411182, | |
| "loss": 3.2907, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 15.64333954354914, | |
| "grad_norm": 0.35753101110458374, | |
| "learning_rate": 0.0004125253348864298, | |
| "loss": 3.294, | |
| "step": 53750 | |
| }, | |
| { | |
| "epoch": 15.657894736842106, | |
| "grad_norm": 0.39829739928245544, | |
| "learning_rate": 0.00041235061153174134, | |
| "loss": 3.2939, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 15.672449930135071, | |
| "grad_norm": 0.36024484038352966, | |
| "learning_rate": 0.00041217588817705293, | |
| "loss": 3.2828, | |
| "step": 53850 | |
| }, | |
| { | |
| "epoch": 15.687005123428039, | |
| "grad_norm": 0.3825265169143677, | |
| "learning_rate": 0.0004120011648223646, | |
| "loss": 3.2886, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 15.701560316721006, | |
| "grad_norm": 0.35621029138565063, | |
| "learning_rate": 0.0004118264414676762, | |
| "loss": 3.2963, | |
| "step": 53950 | |
| }, | |
| { | |
| "epoch": 15.716115510013973, | |
| "grad_norm": 0.36306706070899963, | |
| "learning_rate": 0.0004116517181129877, | |
| "loss": 3.288, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.716115510013973, | |
| "eval_accuracy": 0.372593525585283, | |
| "eval_loss": 3.5393404960632324, | |
| "eval_runtime": 83.3248, | |
| "eval_samples_per_second": 199.833, | |
| "eval_steps_per_second": 12.493, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 15.73067070330694, | |
| "grad_norm": 0.36621716618537903, | |
| "learning_rate": 0.0004114769947582993, | |
| "loss": 3.2993, | |
| "step": 54050 | |
| }, | |
| { | |
| "epoch": 15.745225896599907, | |
| "grad_norm": 0.3718242645263672, | |
| "learning_rate": 0.0004113022714036109, | |
| "loss": 3.3108, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 15.759781089892874, | |
| "grad_norm": 0.3661261796951294, | |
| "learning_rate": 0.0004111275480489225, | |
| "loss": 3.2921, | |
| "step": 54150 | |
| }, | |
| { | |
| "epoch": 15.774336283185841, | |
| "grad_norm": 0.37110430002212524, | |
| "learning_rate": 0.0004109528246942341, | |
| "loss": 3.3036, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 15.788891476478808, | |
| "grad_norm": 0.3758983910083771, | |
| "learning_rate": 0.0004107781013395457, | |
| "loss": 3.2896, | |
| "step": 54250 | |
| }, | |
| { | |
| "epoch": 15.803446669771775, | |
| "grad_norm": 0.36288008093833923, | |
| "learning_rate": 0.0004106033779848573, | |
| "loss": 3.2991, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 15.818001863064742, | |
| "grad_norm": 0.3546515703201294, | |
| "learning_rate": 0.0004104286546301689, | |
| "loss": 3.3009, | |
| "step": 54350 | |
| }, | |
| { | |
| "epoch": 15.83255705635771, | |
| "grad_norm": 0.4252663254737854, | |
| "learning_rate": 0.0004102539312754804, | |
| "loss": 3.3147, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 15.847112249650674, | |
| "grad_norm": 0.3993537724018097, | |
| "learning_rate": 0.00041007920792079206, | |
| "loss": 3.3055, | |
| "step": 54450 | |
| }, | |
| { | |
| "epoch": 15.861667442943642, | |
| "grad_norm": 0.3727772533893585, | |
| "learning_rate": 0.00040990448456610366, | |
| "loss": 3.3112, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 15.876222636236609, | |
| "grad_norm": 0.3480086624622345, | |
| "learning_rate": 0.00040972976121141525, | |
| "loss": 3.309, | |
| "step": 54550 | |
| }, | |
| { | |
| "epoch": 15.890777829529576, | |
| "grad_norm": 0.3650662899017334, | |
| "learning_rate": 0.0004095550378567268, | |
| "loss": 3.2997, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 15.905333022822543, | |
| "grad_norm": 0.374747097492218, | |
| "learning_rate": 0.0004093803145020384, | |
| "loss": 3.3053, | |
| "step": 54650 | |
| }, | |
| { | |
| "epoch": 15.91988821611551, | |
| "grad_norm": 0.3594937026500702, | |
| "learning_rate": 0.00040920559114735, | |
| "loss": 3.3093, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 15.934443409408477, | |
| "grad_norm": 0.358185738325119, | |
| "learning_rate": 0.00040903086779266163, | |
| "loss": 3.3094, | |
| "step": 54750 | |
| }, | |
| { | |
| "epoch": 15.948998602701444, | |
| "grad_norm": 0.3515937924385071, | |
| "learning_rate": 0.00040885614443797317, | |
| "loss": 3.3119, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 15.963553795994411, | |
| "grad_norm": 0.3702413737773895, | |
| "learning_rate": 0.00040868142108328476, | |
| "loss": 3.3058, | |
| "step": 54850 | |
| }, | |
| { | |
| "epoch": 15.978108989287378, | |
| "grad_norm": 0.3735214173793793, | |
| "learning_rate": 0.00040850669772859636, | |
| "loss": 3.3064, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 15.992664182580345, | |
| "grad_norm": 0.3781965970993042, | |
| "learning_rate": 0.0004083319743739079, | |
| "loss": 3.3025, | |
| "step": 54950 | |
| }, | |
| { | |
| "epoch": 16.006986492780623, | |
| "grad_norm": 0.37802356481552124, | |
| "learning_rate": 0.0004081572510192195, | |
| "loss": 3.263, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.006986492780623, | |
| "eval_accuracy": 0.3724095953523122, | |
| "eval_loss": 3.545301675796509, | |
| "eval_runtime": 83.1148, | |
| "eval_samples_per_second": 200.337, | |
| "eval_steps_per_second": 12.525, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 16.02154168607359, | |
| "grad_norm": 0.3845735788345337, | |
| "learning_rate": 0.00040798252766453114, | |
| "loss": 3.2004, | |
| "step": 55050 | |
| }, | |
| { | |
| "epoch": 16.036096879366557, | |
| "grad_norm": 0.36130374670028687, | |
| "learning_rate": 0.00040780780430984273, | |
| "loss": 3.204, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 16.050652072659524, | |
| "grad_norm": 0.3538750410079956, | |
| "learning_rate": 0.0004076330809551543, | |
| "loss": 3.2038, | |
| "step": 55150 | |
| }, | |
| { | |
| "epoch": 16.06520726595249, | |
| "grad_norm": 0.390968918800354, | |
| "learning_rate": 0.00040745835760046587, | |
| "loss": 3.1975, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 16.079762459245458, | |
| "grad_norm": 0.4010736048221588, | |
| "learning_rate": 0.00040728363424577746, | |
| "loss": 3.2023, | |
| "step": 55250 | |
| }, | |
| { | |
| "epoch": 16.094317652538425, | |
| "grad_norm": 0.36839064955711365, | |
| "learning_rate": 0.0004071089108910891, | |
| "loss": 3.2152, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 16.108872845831392, | |
| "grad_norm": 0.354374498128891, | |
| "learning_rate": 0.0004069341875364007, | |
| "loss": 3.2291, | |
| "step": 55350 | |
| }, | |
| { | |
| "epoch": 16.12342803912436, | |
| "grad_norm": 0.3690352439880371, | |
| "learning_rate": 0.00040675946418171225, | |
| "loss": 3.2198, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 16.137983232417326, | |
| "grad_norm": 0.37615934014320374, | |
| "learning_rate": 0.00040658474082702384, | |
| "loss": 3.2235, | |
| "step": 55450 | |
| }, | |
| { | |
| "epoch": 16.152538425710294, | |
| "grad_norm": 0.3690071702003479, | |
| "learning_rate": 0.00040641001747233543, | |
| "loss": 3.2291, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 16.16709361900326, | |
| "grad_norm": 0.38907289505004883, | |
| "learning_rate": 0.000406235294117647, | |
| "loss": 3.2353, | |
| "step": 55550 | |
| }, | |
| { | |
| "epoch": 16.181648812296228, | |
| "grad_norm": 0.37743714451789856, | |
| "learning_rate": 0.0004060605707629586, | |
| "loss": 3.2318, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 16.196204005589195, | |
| "grad_norm": 0.38833099603652954, | |
| "learning_rate": 0.0004058858474082702, | |
| "loss": 3.2402, | |
| "step": 55650 | |
| }, | |
| { | |
| "epoch": 16.21075919888216, | |
| "grad_norm": 0.34554797410964966, | |
| "learning_rate": 0.0004057111240535818, | |
| "loss": 3.2423, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 16.22531439217513, | |
| "grad_norm": 0.39834776520729065, | |
| "learning_rate": 0.00040553640069889335, | |
| "loss": 3.2371, | |
| "step": 55750 | |
| }, | |
| { | |
| "epoch": 16.239869585468096, | |
| "grad_norm": 0.36792176961898804, | |
| "learning_rate": 0.00040536167734420495, | |
| "loss": 3.2505, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 16.254424778761063, | |
| "grad_norm": 0.3809299170970917, | |
| "learning_rate": 0.0004051869539895166, | |
| "loss": 3.2454, | |
| "step": 55850 | |
| }, | |
| { | |
| "epoch": 16.26897997205403, | |
| "grad_norm": 0.365839421749115, | |
| "learning_rate": 0.0004050122306348282, | |
| "loss": 3.2509, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 16.283535165346997, | |
| "grad_norm": 0.3871637284755707, | |
| "learning_rate": 0.00040483750728013973, | |
| "loss": 3.2495, | |
| "step": 55950 | |
| }, | |
| { | |
| "epoch": 16.298090358639964, | |
| "grad_norm": 0.36767056584358215, | |
| "learning_rate": 0.0004046627839254513, | |
| "loss": 3.2338, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.298090358639964, | |
| "eval_accuracy": 0.37286889206186496, | |
| "eval_loss": 3.546008825302124, | |
| "eval_runtime": 82.9947, | |
| "eval_samples_per_second": 200.627, | |
| "eval_steps_per_second": 12.543, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 16.31264555193293, | |
| "grad_norm": 0.3557756841182709, | |
| "learning_rate": 0.0004044880605707629, | |
| "loss": 3.2527, | |
| "step": 56050 | |
| }, | |
| { | |
| "epoch": 16.3272007452259, | |
| "grad_norm": 0.3797302842140198, | |
| "learning_rate": 0.00040431333721607446, | |
| "loss": 3.2451, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 16.341755938518865, | |
| "grad_norm": 0.36076200008392334, | |
| "learning_rate": 0.0004041386138613861, | |
| "loss": 3.2429, | |
| "step": 56150 | |
| }, | |
| { | |
| "epoch": 16.35631113181183, | |
| "grad_norm": 0.36572134494781494, | |
| "learning_rate": 0.0004039638905066977, | |
| "loss": 3.2515, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 16.370866325104796, | |
| "grad_norm": 0.3931342363357544, | |
| "learning_rate": 0.0004037891671520093, | |
| "loss": 3.254, | |
| "step": 56250 | |
| }, | |
| { | |
| "epoch": 16.385421518397763, | |
| "grad_norm": 0.3661358654499054, | |
| "learning_rate": 0.0004036144437973209, | |
| "loss": 3.246, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 16.39997671169073, | |
| "grad_norm": 0.4082087278366089, | |
| "learning_rate": 0.00040343972044263243, | |
| "loss": 3.262, | |
| "step": 56350 | |
| }, | |
| { | |
| "epoch": 16.414531904983697, | |
| "grad_norm": 0.36509233713150024, | |
| "learning_rate": 0.0004032649970879441, | |
| "loss": 3.2625, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 16.429087098276664, | |
| "grad_norm": 0.35747766494750977, | |
| "learning_rate": 0.00040309027373325567, | |
| "loss": 3.2703, | |
| "step": 56450 | |
| }, | |
| { | |
| "epoch": 16.44364229156963, | |
| "grad_norm": 0.36678430438041687, | |
| "learning_rate": 0.00040291555037856727, | |
| "loss": 3.26, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 16.4581974848626, | |
| "grad_norm": 0.39855095744132996, | |
| "learning_rate": 0.0004027408270238788, | |
| "loss": 3.264, | |
| "step": 56550 | |
| }, | |
| { | |
| "epoch": 16.472752678155565, | |
| "grad_norm": 0.39597779512405396, | |
| "learning_rate": 0.0004025661036691904, | |
| "loss": 3.266, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 16.487307871448532, | |
| "grad_norm": 0.3585759401321411, | |
| "learning_rate": 0.000402391380314502, | |
| "loss": 3.2712, | |
| "step": 56650 | |
| }, | |
| { | |
| "epoch": 16.5018630647415, | |
| "grad_norm": 0.39370906352996826, | |
| "learning_rate": 0.00040221665695981364, | |
| "loss": 3.266, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 16.516418258034467, | |
| "grad_norm": 0.3629233241081238, | |
| "learning_rate": 0.0004020419336051252, | |
| "loss": 3.2685, | |
| "step": 56750 | |
| }, | |
| { | |
| "epoch": 16.530973451327434, | |
| "grad_norm": 0.38627785444259644, | |
| "learning_rate": 0.0004018672102504368, | |
| "loss": 3.2704, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 16.5455286446204, | |
| "grad_norm": 0.36387115716934204, | |
| "learning_rate": 0.00040169248689574837, | |
| "loss": 3.2757, | |
| "step": 56850 | |
| }, | |
| { | |
| "epoch": 16.560083837913368, | |
| "grad_norm": 0.403450071811676, | |
| "learning_rate": 0.0004015177635410599, | |
| "loss": 3.2822, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 16.574639031206335, | |
| "grad_norm": 0.3927227854728699, | |
| "learning_rate": 0.0004013430401863715, | |
| "loss": 3.2597, | |
| "step": 56950 | |
| }, | |
| { | |
| "epoch": 16.589194224499302, | |
| "grad_norm": 0.36201542615890503, | |
| "learning_rate": 0.00040116831683168315, | |
| "loss": 3.2846, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.589194224499302, | |
| "eval_accuracy": 0.37280483968041506, | |
| "eval_loss": 3.5414578914642334, | |
| "eval_runtime": 83.657, | |
| "eval_samples_per_second": 199.039, | |
| "eval_steps_per_second": 12.444, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 16.60374941779227, | |
| "grad_norm": 0.34870487451553345, | |
| "learning_rate": 0.00040099359347699475, | |
| "loss": 3.2786, | |
| "step": 57050 | |
| }, | |
| { | |
| "epoch": 16.618304611085236, | |
| "grad_norm": 0.35893914103507996, | |
| "learning_rate": 0.0004008188701223063, | |
| "loss": 3.2728, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 16.632859804378203, | |
| "grad_norm": 0.37824079394340515, | |
| "learning_rate": 0.0004006441467676179, | |
| "loss": 3.2815, | |
| "step": 57150 | |
| }, | |
| { | |
| "epoch": 16.64741499767117, | |
| "grad_norm": 0.36958247423171997, | |
| "learning_rate": 0.0004004694234129295, | |
| "loss": 3.2852, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 16.661970190964137, | |
| "grad_norm": 0.35305625200271606, | |
| "learning_rate": 0.0004002947000582411, | |
| "loss": 3.2663, | |
| "step": 57250 | |
| }, | |
| { | |
| "epoch": 16.676525384257104, | |
| "grad_norm": 0.3525847792625427, | |
| "learning_rate": 0.00040011997670355267, | |
| "loss": 3.2789, | |
| "step": 57300 | |
| }, | |
| { | |
| "epoch": 16.69108057755007, | |
| "grad_norm": 0.3740130364894867, | |
| "learning_rate": 0.00039994525334886426, | |
| "loss": 3.2876, | |
| "step": 57350 | |
| }, | |
| { | |
| "epoch": 16.70563577084304, | |
| "grad_norm": 0.38981494307518005, | |
| "learning_rate": 0.00039977052999417585, | |
| "loss": 3.2809, | |
| "step": 57400 | |
| }, | |
| { | |
| "epoch": 16.720190964136005, | |
| "grad_norm": 0.3750765323638916, | |
| "learning_rate": 0.00039959580663948745, | |
| "loss": 3.2902, | |
| "step": 57450 | |
| }, | |
| { | |
| "epoch": 16.734746157428972, | |
| "grad_norm": 0.3791131377220154, | |
| "learning_rate": 0.000399421083284799, | |
| "loss": 3.2631, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 16.749301350721936, | |
| "grad_norm": 0.3747536242008209, | |
| "learning_rate": 0.00039924635993011064, | |
| "loss": 3.2893, | |
| "step": 57550 | |
| }, | |
| { | |
| "epoch": 16.763856544014903, | |
| "grad_norm": 0.39232292771339417, | |
| "learning_rate": 0.00039907163657542223, | |
| "loss": 3.3006, | |
| "step": 57600 | |
| }, | |
| { | |
| "epoch": 16.77841173730787, | |
| "grad_norm": 0.38699984550476074, | |
| "learning_rate": 0.0003988969132207338, | |
| "loss": 3.2866, | |
| "step": 57650 | |
| }, | |
| { | |
| "epoch": 16.792966930600837, | |
| "grad_norm": 0.35927581787109375, | |
| "learning_rate": 0.00039872218986604537, | |
| "loss": 3.2714, | |
| "step": 57700 | |
| }, | |
| { | |
| "epoch": 16.807522123893804, | |
| "grad_norm": 0.3530392050743103, | |
| "learning_rate": 0.00039854746651135696, | |
| "loss": 3.2725, | |
| "step": 57750 | |
| }, | |
| { | |
| "epoch": 16.82207731718677, | |
| "grad_norm": 0.37735065817832947, | |
| "learning_rate": 0.0003983727431566686, | |
| "loss": 3.2847, | |
| "step": 57800 | |
| }, | |
| { | |
| "epoch": 16.83663251047974, | |
| "grad_norm": 0.38122105598449707, | |
| "learning_rate": 0.0003981980198019802, | |
| "loss": 3.2763, | |
| "step": 57850 | |
| }, | |
| { | |
| "epoch": 16.851187703772705, | |
| "grad_norm": 0.3847242295742035, | |
| "learning_rate": 0.00039802329644729174, | |
| "loss": 3.2879, | |
| "step": 57900 | |
| }, | |
| { | |
| "epoch": 16.865742897065672, | |
| "grad_norm": 0.36237791180610657, | |
| "learning_rate": 0.00039784857309260334, | |
| "loss": 3.2785, | |
| "step": 57950 | |
| }, | |
| { | |
| "epoch": 16.88029809035864, | |
| "grad_norm": 0.37167662382125854, | |
| "learning_rate": 0.00039767384973791493, | |
| "loss": 3.2937, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.88029809035864, | |
| "eval_accuracy": 0.3733340651366884, | |
| "eval_loss": 3.5342485904693604, | |
| "eval_runtime": 83.7234, | |
| "eval_samples_per_second": 198.881, | |
| "eval_steps_per_second": 12.434, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 16.894853283651607, | |
| "grad_norm": 0.37969404458999634, | |
| "learning_rate": 0.00039749912638322647, | |
| "loss": 3.2868, | |
| "step": 58050 | |
| }, | |
| { | |
| "epoch": 16.909408476944574, | |
| "grad_norm": 0.3962460160255432, | |
| "learning_rate": 0.0003973244030285381, | |
| "loss": 3.2898, | |
| "step": 58100 | |
| }, | |
| { | |
| "epoch": 16.92396367023754, | |
| "grad_norm": 0.4090690612792969, | |
| "learning_rate": 0.0003971496796738497, | |
| "loss": 3.2927, | |
| "step": 58150 | |
| }, | |
| { | |
| "epoch": 16.938518863530508, | |
| "grad_norm": 0.37081751227378845, | |
| "learning_rate": 0.0003969749563191613, | |
| "loss": 3.2939, | |
| "step": 58200 | |
| }, | |
| { | |
| "epoch": 16.953074056823475, | |
| "grad_norm": 0.3689468204975128, | |
| "learning_rate": 0.00039680023296447285, | |
| "loss": 3.286, | |
| "step": 58250 | |
| }, | |
| { | |
| "epoch": 16.967629250116442, | |
| "grad_norm": 0.4149116277694702, | |
| "learning_rate": 0.00039662550960978444, | |
| "loss": 3.2745, | |
| "step": 58300 | |
| }, | |
| { | |
| "epoch": 16.98218444340941, | |
| "grad_norm": 0.363203763961792, | |
| "learning_rate": 0.00039645078625509604, | |
| "loss": 3.2905, | |
| "step": 58350 | |
| }, | |
| { | |
| "epoch": 16.996739636702376, | |
| "grad_norm": 0.3528755307197571, | |
| "learning_rate": 0.0003962760629004077, | |
| "loss": 3.2923, | |
| "step": 58400 | |
| }, | |
| { | |
| "epoch": 17.011061946902654, | |
| "grad_norm": 0.389476478099823, | |
| "learning_rate": 0.0003961013395457193, | |
| "loss": 3.1959, | |
| "step": 58450 | |
| }, | |
| { | |
| "epoch": 17.02561714019562, | |
| "grad_norm": 0.35332250595092773, | |
| "learning_rate": 0.0003959266161910308, | |
| "loss": 3.1849, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 17.040172333488588, | |
| "grad_norm": 0.3894152343273163, | |
| "learning_rate": 0.0003957518928363424, | |
| "loss": 3.1926, | |
| "step": 58550 | |
| }, | |
| { | |
| "epoch": 17.054727526781555, | |
| "grad_norm": 0.3673129677772522, | |
| "learning_rate": 0.000395577169481654, | |
| "loss": 3.1967, | |
| "step": 58600 | |
| }, | |
| { | |
| "epoch": 17.069282720074522, | |
| "grad_norm": 0.3771451413631439, | |
| "learning_rate": 0.00039540244612696566, | |
| "loss": 3.2037, | |
| "step": 58650 | |
| }, | |
| { | |
| "epoch": 17.08383791336749, | |
| "grad_norm": 0.371507465839386, | |
| "learning_rate": 0.0003952277227722772, | |
| "loss": 3.2054, | |
| "step": 58700 | |
| }, | |
| { | |
| "epoch": 17.098393106660456, | |
| "grad_norm": 0.374114066362381, | |
| "learning_rate": 0.0003950529994175888, | |
| "loss": 3.192, | |
| "step": 58750 | |
| }, | |
| { | |
| "epoch": 17.112948299953423, | |
| "grad_norm": 0.3767269253730774, | |
| "learning_rate": 0.0003948782760629004, | |
| "loss": 3.2087, | |
| "step": 58800 | |
| }, | |
| { | |
| "epoch": 17.12750349324639, | |
| "grad_norm": 0.3896580636501312, | |
| "learning_rate": 0.0003947035527082119, | |
| "loss": 3.2207, | |
| "step": 58850 | |
| }, | |
| { | |
| "epoch": 17.142058686539357, | |
| "grad_norm": 0.39091426134109497, | |
| "learning_rate": 0.0003945288293535235, | |
| "loss": 3.2112, | |
| "step": 58900 | |
| }, | |
| { | |
| "epoch": 17.156613879832324, | |
| "grad_norm": 0.4273008406162262, | |
| "learning_rate": 0.00039435410599883517, | |
| "loss": 3.2206, | |
| "step": 58950 | |
| }, | |
| { | |
| "epoch": 17.17116907312529, | |
| "grad_norm": 0.37491142749786377, | |
| "learning_rate": 0.00039417938264414676, | |
| "loss": 3.2176, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.17116907312529, | |
| "eval_accuracy": 0.37257448616180616, | |
| "eval_loss": 3.549081563949585, | |
| "eval_runtime": 83.8013, | |
| "eval_samples_per_second": 198.696, | |
| "eval_steps_per_second": 12.422, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 17.18572426641826, | |
| "grad_norm": 0.36968278884887695, | |
| "learning_rate": 0.0003940046592894583, | |
| "loss": 3.2228, | |
| "step": 59050 | |
| }, | |
| { | |
| "epoch": 17.200279459711226, | |
| "grad_norm": 0.41922426223754883, | |
| "learning_rate": 0.0003938299359347699, | |
| "loss": 3.2225, | |
| "step": 59100 | |
| }, | |
| { | |
| "epoch": 17.214834653004193, | |
| "grad_norm": 0.36802321672439575, | |
| "learning_rate": 0.0003936552125800815, | |
| "loss": 3.2283, | |
| "step": 59150 | |
| }, | |
| { | |
| "epoch": 17.22938984629716, | |
| "grad_norm": 0.3892861306667328, | |
| "learning_rate": 0.00039348048922539314, | |
| "loss": 3.2124, | |
| "step": 59200 | |
| }, | |
| { | |
| "epoch": 17.243945039590127, | |
| "grad_norm": 0.37343189120292664, | |
| "learning_rate": 0.0003933057658707047, | |
| "loss": 3.2312, | |
| "step": 59250 | |
| }, | |
| { | |
| "epoch": 17.258500232883094, | |
| "grad_norm": 0.39486169815063477, | |
| "learning_rate": 0.0003931310425160163, | |
| "loss": 3.225, | |
| "step": 59300 | |
| }, | |
| { | |
| "epoch": 17.27305542617606, | |
| "grad_norm": 0.3588271141052246, | |
| "learning_rate": 0.00039295631916132787, | |
| "loss": 3.2316, | |
| "step": 59350 | |
| }, | |
| { | |
| "epoch": 17.287610619469028, | |
| "grad_norm": 0.3605259656906128, | |
| "learning_rate": 0.00039278159580663946, | |
| "loss": 3.2233, | |
| "step": 59400 | |
| }, | |
| { | |
| "epoch": 17.302165812761995, | |
| "grad_norm": 0.3496353030204773, | |
| "learning_rate": 0.000392606872451951, | |
| "loss": 3.2386, | |
| "step": 59450 | |
| }, | |
| { | |
| "epoch": 17.316721006054962, | |
| "grad_norm": 0.3507852852344513, | |
| "learning_rate": 0.00039243214909726265, | |
| "loss": 3.2316, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 17.331276199347926, | |
| "grad_norm": 0.3779459297657013, | |
| "learning_rate": 0.00039225742574257425, | |
| "loss": 3.238, | |
| "step": 59550 | |
| }, | |
| { | |
| "epoch": 17.345831392640893, | |
| "grad_norm": 0.3823668956756592, | |
| "learning_rate": 0.00039208270238788584, | |
| "loss": 3.2421, | |
| "step": 59600 | |
| }, | |
| { | |
| "epoch": 17.36038658593386, | |
| "grad_norm": 0.40699446201324463, | |
| "learning_rate": 0.0003919079790331974, | |
| "loss": 3.2382, | |
| "step": 59650 | |
| }, | |
| { | |
| "epoch": 17.374941779226827, | |
| "grad_norm": 0.39647895097732544, | |
| "learning_rate": 0.000391733255678509, | |
| "loss": 3.2351, | |
| "step": 59700 | |
| }, | |
| { | |
| "epoch": 17.389496972519794, | |
| "grad_norm": 0.39623236656188965, | |
| "learning_rate": 0.00039155853232382057, | |
| "loss": 3.2469, | |
| "step": 59750 | |
| }, | |
| { | |
| "epoch": 17.40405216581276, | |
| "grad_norm": 0.37540996074676514, | |
| "learning_rate": 0.0003913838089691322, | |
| "loss": 3.2423, | |
| "step": 59800 | |
| }, | |
| { | |
| "epoch": 17.418607359105728, | |
| "grad_norm": 0.3802138566970825, | |
| "learning_rate": 0.00039120908561444376, | |
| "loss": 3.2527, | |
| "step": 59850 | |
| }, | |
| { | |
| "epoch": 17.433162552398695, | |
| "grad_norm": 0.3569558262825012, | |
| "learning_rate": 0.00039103436225975535, | |
| "loss": 3.2527, | |
| "step": 59900 | |
| }, | |
| { | |
| "epoch": 17.447717745691662, | |
| "grad_norm": 0.3840680718421936, | |
| "learning_rate": 0.00039085963890506695, | |
| "loss": 3.2521, | |
| "step": 59950 | |
| }, | |
| { | |
| "epoch": 17.46227293898463, | |
| "grad_norm": 0.3617042005062103, | |
| "learning_rate": 0.0003906849155503785, | |
| "loss": 3.2612, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.46227293898463, | |
| "eval_accuracy": 0.37305141196717084, | |
| "eval_loss": 3.543017864227295, | |
| "eval_runtime": 83.9502, | |
| "eval_samples_per_second": 198.344, | |
| "eval_steps_per_second": 12.4, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 17.476828132277596, | |
| "grad_norm": 0.38519003987312317, | |
| "learning_rate": 0.00039051019219569014, | |
| "loss": 3.2453, | |
| "step": 60050 | |
| }, | |
| { | |
| "epoch": 17.491383325570563, | |
| "grad_norm": 0.39966607093811035, | |
| "learning_rate": 0.00039033546884100173, | |
| "loss": 3.2405, | |
| "step": 60100 | |
| }, | |
| { | |
| "epoch": 17.50593851886353, | |
| "grad_norm": 0.37919265031814575, | |
| "learning_rate": 0.0003901607454863133, | |
| "loss": 3.2642, | |
| "step": 60150 | |
| }, | |
| { | |
| "epoch": 17.520493712156497, | |
| "grad_norm": 0.3577299118041992, | |
| "learning_rate": 0.00038998602213162486, | |
| "loss": 3.2534, | |
| "step": 60200 | |
| }, | |
| { | |
| "epoch": 17.535048905449464, | |
| "grad_norm": 0.37814420461654663, | |
| "learning_rate": 0.00038981129877693646, | |
| "loss": 3.251, | |
| "step": 60250 | |
| }, | |
| { | |
| "epoch": 17.54960409874243, | |
| "grad_norm": 0.374645471572876, | |
| "learning_rate": 0.00038963657542224805, | |
| "loss": 3.258, | |
| "step": 60300 | |
| }, | |
| { | |
| "epoch": 17.5641592920354, | |
| "grad_norm": 0.3777361810207367, | |
| "learning_rate": 0.0003894618520675597, | |
| "loss": 3.2611, | |
| "step": 60350 | |
| }, | |
| { | |
| "epoch": 17.578714485328366, | |
| "grad_norm": 0.3927379846572876, | |
| "learning_rate": 0.00038928712871287124, | |
| "loss": 3.2486, | |
| "step": 60400 | |
| }, | |
| { | |
| "epoch": 17.593269678621333, | |
| "grad_norm": 0.38195985555648804, | |
| "learning_rate": 0.00038911240535818284, | |
| "loss": 3.2738, | |
| "step": 60450 | |
| }, | |
| { | |
| "epoch": 17.6078248719143, | |
| "grad_norm": 0.3771248757839203, | |
| "learning_rate": 0.00038893768200349443, | |
| "loss": 3.263, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 17.622380065207267, | |
| "grad_norm": 0.3958074450492859, | |
| "learning_rate": 0.000388762958648806, | |
| "loss": 3.2584, | |
| "step": 60550 | |
| }, | |
| { | |
| "epoch": 17.636935258500234, | |
| "grad_norm": 0.4122246205806732, | |
| "learning_rate": 0.0003885882352941176, | |
| "loss": 3.2752, | |
| "step": 60600 | |
| }, | |
| { | |
| "epoch": 17.6514904517932, | |
| "grad_norm": 0.36372998356819153, | |
| "learning_rate": 0.0003884135119394292, | |
| "loss": 3.2568, | |
| "step": 60650 | |
| }, | |
| { | |
| "epoch": 17.666045645086168, | |
| "grad_norm": 0.36799973249435425, | |
| "learning_rate": 0.0003882387885847408, | |
| "loss": 3.2616, | |
| "step": 60700 | |
| }, | |
| { | |
| "epoch": 17.680600838379135, | |
| "grad_norm": 0.3644178509712219, | |
| "learning_rate": 0.0003880640652300524, | |
| "loss": 3.2629, | |
| "step": 60750 | |
| }, | |
| { | |
| "epoch": 17.695156031672102, | |
| "grad_norm": 0.3425600826740265, | |
| "learning_rate": 0.00038788934187536394, | |
| "loss": 3.2702, | |
| "step": 60800 | |
| }, | |
| { | |
| "epoch": 17.70971122496507, | |
| "grad_norm": 0.36027956008911133, | |
| "learning_rate": 0.00038771461852067554, | |
| "loss": 3.2746, | |
| "step": 60850 | |
| }, | |
| { | |
| "epoch": 17.724266418258033, | |
| "grad_norm": 0.35802096128463745, | |
| "learning_rate": 0.0003875398951659872, | |
| "loss": 3.2718, | |
| "step": 60900 | |
| }, | |
| { | |
| "epoch": 17.738821611551, | |
| "grad_norm": 0.3740179240703583, | |
| "learning_rate": 0.0003873651718112988, | |
| "loss": 3.2727, | |
| "step": 60950 | |
| }, | |
| { | |
| "epoch": 17.753376804843967, | |
| "grad_norm": 0.3695901334285736, | |
| "learning_rate": 0.0003871904484566103, | |
| "loss": 3.2734, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.753376804843967, | |
| "eval_accuracy": 0.37353574199277656, | |
| "eval_loss": 3.5319979190826416, | |
| "eval_runtime": 84.0386, | |
| "eval_samples_per_second": 198.135, | |
| "eval_steps_per_second": 12.387, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 17.767931998136934, | |
| "grad_norm": 0.35077133774757385, | |
| "learning_rate": 0.0003870157251019219, | |
| "loss": 3.2721, | |
| "step": 61050 | |
| }, | |
| { | |
| "epoch": 17.7824871914299, | |
| "grad_norm": 0.37661030888557434, | |
| "learning_rate": 0.0003868410017472335, | |
| "loss": 3.2766, | |
| "step": 61100 | |
| }, | |
| { | |
| "epoch": 17.797042384722868, | |
| "grad_norm": 0.36504772305488586, | |
| "learning_rate": 0.00038666627839254505, | |
| "loss": 3.2678, | |
| "step": 61150 | |
| }, | |
| { | |
| "epoch": 17.811597578015835, | |
| "grad_norm": 0.3955831527709961, | |
| "learning_rate": 0.0003864915550378567, | |
| "loss": 3.2777, | |
| "step": 61200 | |
| }, | |
| { | |
| "epoch": 17.826152771308802, | |
| "grad_norm": 0.37011846899986267, | |
| "learning_rate": 0.0003863168316831683, | |
| "loss": 3.2819, | |
| "step": 61250 | |
| }, | |
| { | |
| "epoch": 17.84070796460177, | |
| "grad_norm": 0.3678891658782959, | |
| "learning_rate": 0.0003861421083284799, | |
| "loss": 3.2849, | |
| "step": 61300 | |
| }, | |
| { | |
| "epoch": 17.855263157894736, | |
| "grad_norm": 0.38564667105674744, | |
| "learning_rate": 0.0003859673849737914, | |
| "loss": 3.2725, | |
| "step": 61350 | |
| }, | |
| { | |
| "epoch": 17.869818351187703, | |
| "grad_norm": 0.38152021169662476, | |
| "learning_rate": 0.000385792661619103, | |
| "loss": 3.2839, | |
| "step": 61400 | |
| }, | |
| { | |
| "epoch": 17.88437354448067, | |
| "grad_norm": 0.3962383568286896, | |
| "learning_rate": 0.00038561793826441467, | |
| "loss": 3.2751, | |
| "step": 61450 | |
| }, | |
| { | |
| "epoch": 17.898928737773637, | |
| "grad_norm": 0.380815714597702, | |
| "learning_rate": 0.00038544321490972626, | |
| "loss": 3.2872, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 17.913483931066605, | |
| "grad_norm": 0.3607582449913025, | |
| "learning_rate": 0.0003852684915550378, | |
| "loss": 3.2707, | |
| "step": 61550 | |
| }, | |
| { | |
| "epoch": 17.92803912435957, | |
| "grad_norm": 0.3585313856601715, | |
| "learning_rate": 0.0003850937682003494, | |
| "loss": 3.2731, | |
| "step": 61600 | |
| }, | |
| { | |
| "epoch": 17.94259431765254, | |
| "grad_norm": 0.38443124294281006, | |
| "learning_rate": 0.000384919044845661, | |
| "loss": 3.2859, | |
| "step": 61650 | |
| }, | |
| { | |
| "epoch": 17.957149510945506, | |
| "grad_norm": 0.38093116879463196, | |
| "learning_rate": 0.0003847443214909726, | |
| "loss": 3.2647, | |
| "step": 61700 | |
| }, | |
| { | |
| "epoch": 17.971704704238473, | |
| "grad_norm": 0.3577301800251007, | |
| "learning_rate": 0.00038456959813628423, | |
| "loss": 3.2784, | |
| "step": 61750 | |
| }, | |
| { | |
| "epoch": 17.98625989753144, | |
| "grad_norm": 0.3698591887950897, | |
| "learning_rate": 0.0003843948747815958, | |
| "loss": 3.289, | |
| "step": 61800 | |
| }, | |
| { | |
| "epoch": 18.000582207731718, | |
| "grad_norm": 0.42908018827438354, | |
| "learning_rate": 0.00038422015142690737, | |
| "loss": 3.2847, | |
| "step": 61850 | |
| }, | |
| { | |
| "epoch": 18.015137401024685, | |
| "grad_norm": 0.3832159638404846, | |
| "learning_rate": 0.00038404542807221896, | |
| "loss": 3.1655, | |
| "step": 61900 | |
| }, | |
| { | |
| "epoch": 18.029692594317652, | |
| "grad_norm": 0.4008435904979706, | |
| "learning_rate": 0.0003838707047175305, | |
| "loss": 3.163, | |
| "step": 61950 | |
| }, | |
| { | |
| "epoch": 18.04424778761062, | |
| "grad_norm": 0.3579607903957367, | |
| "learning_rate": 0.00038369598136284215, | |
| "loss": 3.171, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.04424778761062, | |
| "eval_accuracy": 0.3730583460781902, | |
| "eval_loss": 3.5451226234436035, | |
| "eval_runtime": 84.0663, | |
| "eval_samples_per_second": 198.07, | |
| "eval_steps_per_second": 12.383, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 18.058802980903586, | |
| "grad_norm": 0.3776082992553711, | |
| "learning_rate": 0.00038352125800815374, | |
| "loss": 3.1763, | |
| "step": 62050 | |
| }, | |
| { | |
| "epoch": 18.073358174196553, | |
| "grad_norm": 0.3655526340007782, | |
| "learning_rate": 0.00038334653465346534, | |
| "loss": 3.1838, | |
| "step": 62100 | |
| }, | |
| { | |
| "epoch": 18.08791336748952, | |
| "grad_norm": 0.38094913959503174, | |
| "learning_rate": 0.0003831718112987769, | |
| "loss": 3.1917, | |
| "step": 62150 | |
| }, | |
| { | |
| "epoch": 18.102468560782487, | |
| "grad_norm": 0.3664492070674896, | |
| "learning_rate": 0.0003829970879440885, | |
| "loss": 3.1952, | |
| "step": 62200 | |
| }, | |
| { | |
| "epoch": 18.117023754075454, | |
| "grad_norm": 0.38657721877098083, | |
| "learning_rate": 0.00038282236458940007, | |
| "loss": 3.2111, | |
| "step": 62250 | |
| }, | |
| { | |
| "epoch": 18.13157894736842, | |
| "grad_norm": 0.37256214022636414, | |
| "learning_rate": 0.0003826476412347117, | |
| "loss": 3.181, | |
| "step": 62300 | |
| }, | |
| { | |
| "epoch": 18.14613414066139, | |
| "grad_norm": 0.37336111068725586, | |
| "learning_rate": 0.00038247291788002326, | |
| "loss": 3.2069, | |
| "step": 62350 | |
| }, | |
| { | |
| "epoch": 18.160689333954355, | |
| "grad_norm": 0.41406548023223877, | |
| "learning_rate": 0.00038229819452533485, | |
| "loss": 3.2146, | |
| "step": 62400 | |
| }, | |
| { | |
| "epoch": 18.175244527247322, | |
| "grad_norm": 0.3549617528915405, | |
| "learning_rate": 0.00038212347117064644, | |
| "loss": 3.2133, | |
| "step": 62450 | |
| }, | |
| { | |
| "epoch": 18.18979972054029, | |
| "grad_norm": 0.3977498412132263, | |
| "learning_rate": 0.000381948747815958, | |
| "loss": 3.2057, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 18.204354913833257, | |
| "grad_norm": 0.38086074590682983, | |
| "learning_rate": 0.0003817740244612696, | |
| "loss": 3.2108, | |
| "step": 62550 | |
| }, | |
| { | |
| "epoch": 18.218910107126224, | |
| "grad_norm": 0.3952549397945404, | |
| "learning_rate": 0.00038159930110658123, | |
| "loss": 3.2153, | |
| "step": 62600 | |
| }, | |
| { | |
| "epoch": 18.23346530041919, | |
| "grad_norm": 0.3730448782444, | |
| "learning_rate": 0.0003814245777518928, | |
| "loss": 3.2104, | |
| "step": 62650 | |
| }, | |
| { | |
| "epoch": 18.248020493712158, | |
| "grad_norm": 0.40390902757644653, | |
| "learning_rate": 0.0003812498543972044, | |
| "loss": 3.2189, | |
| "step": 62700 | |
| }, | |
| { | |
| "epoch": 18.262575687005125, | |
| "grad_norm": 0.3541817367076874, | |
| "learning_rate": 0.00038107513104251596, | |
| "loss": 3.2261, | |
| "step": 62750 | |
| }, | |
| { | |
| "epoch": 18.277130880298092, | |
| "grad_norm": 0.38586553931236267, | |
| "learning_rate": 0.00038090040768782755, | |
| "loss": 3.231, | |
| "step": 62800 | |
| }, | |
| { | |
| "epoch": 18.29168607359106, | |
| "grad_norm": 0.37818339467048645, | |
| "learning_rate": 0.0003807256843331392, | |
| "loss": 3.233, | |
| "step": 62850 | |
| }, | |
| { | |
| "epoch": 18.306241266884022, | |
| "grad_norm": 0.39146703481674194, | |
| "learning_rate": 0.0003805509609784508, | |
| "loss": 3.2244, | |
| "step": 62900 | |
| }, | |
| { | |
| "epoch": 18.32079646017699, | |
| "grad_norm": 0.36770719289779663, | |
| "learning_rate": 0.00038037623762376233, | |
| "loss": 3.2285, | |
| "step": 62950 | |
| }, | |
| { | |
| "epoch": 18.335351653469957, | |
| "grad_norm": 0.375123530626297, | |
| "learning_rate": 0.00038020151426907393, | |
| "loss": 3.2308, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.335351653469957, | |
| "eval_accuracy": 0.3729977019885973, | |
| "eval_loss": 3.5478463172912598, | |
| "eval_runtime": 84.0176, | |
| "eval_samples_per_second": 198.185, | |
| "eval_steps_per_second": 12.39, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 18.349906846762924, | |
| "grad_norm": 0.3701789677143097, | |
| "learning_rate": 0.0003800267909143855, | |
| "loss": 3.2247, | |
| "step": 63050 | |
| }, | |
| { | |
| "epoch": 18.36446204005589, | |
| "grad_norm": 0.37322163581848145, | |
| "learning_rate": 0.00037985206755969706, | |
| "loss": 3.2285, | |
| "step": 63100 | |
| }, | |
| { | |
| "epoch": 18.379017233348858, | |
| "grad_norm": 0.38248884677886963, | |
| "learning_rate": 0.0003796773442050087, | |
| "loss": 3.2268, | |
| "step": 63150 | |
| }, | |
| { | |
| "epoch": 18.393572426641825, | |
| "grad_norm": 0.38328638672828674, | |
| "learning_rate": 0.0003795026208503203, | |
| "loss": 3.2488, | |
| "step": 63200 | |
| }, | |
| { | |
| "epoch": 18.408127619934792, | |
| "grad_norm": 0.3659778833389282, | |
| "learning_rate": 0.0003793278974956319, | |
| "loss": 3.2569, | |
| "step": 63250 | |
| }, | |
| { | |
| "epoch": 18.42268281322776, | |
| "grad_norm": 0.38923901319503784, | |
| "learning_rate": 0.00037915317414094344, | |
| "loss": 3.2478, | |
| "step": 63300 | |
| }, | |
| { | |
| "epoch": 18.437238006520726, | |
| "grad_norm": 0.37889325618743896, | |
| "learning_rate": 0.00037897845078625503, | |
| "loss": 3.2373, | |
| "step": 63350 | |
| }, | |
| { | |
| "epoch": 18.451793199813693, | |
| "grad_norm": 0.38472655415534973, | |
| "learning_rate": 0.0003788037274315667, | |
| "loss": 3.244, | |
| "step": 63400 | |
| }, | |
| { | |
| "epoch": 18.46634839310666, | |
| "grad_norm": 0.3745638430118561, | |
| "learning_rate": 0.0003786290040768783, | |
| "loss": 3.2449, | |
| "step": 63450 | |
| }, | |
| { | |
| "epoch": 18.480903586399627, | |
| "grad_norm": 0.39063334465026855, | |
| "learning_rate": 0.0003784542807221898, | |
| "loss": 3.2287, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 18.495458779692594, | |
| "grad_norm": 0.3969773054122925, | |
| "learning_rate": 0.0003782795573675014, | |
| "loss": 3.2357, | |
| "step": 63550 | |
| }, | |
| { | |
| "epoch": 18.51001397298556, | |
| "grad_norm": 0.3670395314693451, | |
| "learning_rate": 0.000378104834012813, | |
| "loss": 3.2619, | |
| "step": 63600 | |
| }, | |
| { | |
| "epoch": 18.52456916627853, | |
| "grad_norm": 0.39414724707603455, | |
| "learning_rate": 0.0003779301106581246, | |
| "loss": 3.2416, | |
| "step": 63650 | |
| }, | |
| { | |
| "epoch": 18.539124359571495, | |
| "grad_norm": 0.3680049180984497, | |
| "learning_rate": 0.0003777553873034362, | |
| "loss": 3.2527, | |
| "step": 63700 | |
| }, | |
| { | |
| "epoch": 18.553679552864462, | |
| "grad_norm": 0.3810567259788513, | |
| "learning_rate": 0.0003775806639487478, | |
| "loss": 3.2477, | |
| "step": 63750 | |
| }, | |
| { | |
| "epoch": 18.56823474615743, | |
| "grad_norm": 0.38676631450653076, | |
| "learning_rate": 0.0003774059405940594, | |
| "loss": 3.2352, | |
| "step": 63800 | |
| }, | |
| { | |
| "epoch": 18.582789939450397, | |
| "grad_norm": 0.38043132424354553, | |
| "learning_rate": 0.000377231217239371, | |
| "loss": 3.24, | |
| "step": 63850 | |
| }, | |
| { | |
| "epoch": 18.597345132743364, | |
| "grad_norm": 0.3772090673446655, | |
| "learning_rate": 0.0003770564938846825, | |
| "loss": 3.2398, | |
| "step": 63900 | |
| }, | |
| { | |
| "epoch": 18.61190032603633, | |
| "grad_norm": 0.40552330017089844, | |
| "learning_rate": 0.00037688177052999416, | |
| "loss": 3.2501, | |
| "step": 63950 | |
| }, | |
| { | |
| "epoch": 18.626455519329298, | |
| "grad_norm": 0.35742637515068054, | |
| "learning_rate": 0.00037670704717530576, | |
| "loss": 3.2547, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.626455519329298, | |
| "eval_accuracy": 0.3734908465621089, | |
| "eval_loss": 3.537060260772705, | |
| "eval_runtime": 84.054, | |
| "eval_samples_per_second": 198.099, | |
| "eval_steps_per_second": 12.385, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 18.641010712622265, | |
| "grad_norm": 0.4211755096912384, | |
| "learning_rate": 0.00037653232382061735, | |
| "loss": 3.2524, | |
| "step": 64050 | |
| }, | |
| { | |
| "epoch": 18.655565905915232, | |
| "grad_norm": 0.37635040283203125, | |
| "learning_rate": 0.0003763576004659289, | |
| "loss": 3.2498, | |
| "step": 64100 | |
| }, | |
| { | |
| "epoch": 18.6701210992082, | |
| "grad_norm": 0.3736768066883087, | |
| "learning_rate": 0.0003761828771112405, | |
| "loss": 3.253, | |
| "step": 64150 | |
| }, | |
| { | |
| "epoch": 18.684676292501166, | |
| "grad_norm": 0.37573927640914917, | |
| "learning_rate": 0.0003760081537565521, | |
| "loss": 3.2509, | |
| "step": 64200 | |
| }, | |
| { | |
| "epoch": 18.69923148579413, | |
| "grad_norm": 0.3648873269557953, | |
| "learning_rate": 0.00037583343040186373, | |
| "loss": 3.2463, | |
| "step": 64250 | |
| }, | |
| { | |
| "epoch": 18.713786679087097, | |
| "grad_norm": 0.4201739728450775, | |
| "learning_rate": 0.00037565870704717527, | |
| "loss": 3.2636, | |
| "step": 64300 | |
| }, | |
| { | |
| "epoch": 18.728341872380064, | |
| "grad_norm": 0.37984031438827515, | |
| "learning_rate": 0.00037548398369248687, | |
| "loss": 3.2592, | |
| "step": 64350 | |
| }, | |
| { | |
| "epoch": 18.74289706567303, | |
| "grad_norm": 0.3736239969730377, | |
| "learning_rate": 0.00037530926033779846, | |
| "loss": 3.2693, | |
| "step": 64400 | |
| }, | |
| { | |
| "epoch": 18.757452258965998, | |
| "grad_norm": 0.3848115801811218, | |
| "learning_rate": 0.00037513453698311, | |
| "loss": 3.2642, | |
| "step": 64450 | |
| }, | |
| { | |
| "epoch": 18.772007452258965, | |
| "grad_norm": 0.39798057079315186, | |
| "learning_rate": 0.0003749598136284216, | |
| "loss": 3.2582, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 18.786562645551932, | |
| "grad_norm": 0.3883126974105835, | |
| "learning_rate": 0.00037478509027373324, | |
| "loss": 3.2627, | |
| "step": 64550 | |
| }, | |
| { | |
| "epoch": 18.8011178388449, | |
| "grad_norm": 0.38083627820014954, | |
| "learning_rate": 0.00037461036691904484, | |
| "loss": 3.2443, | |
| "step": 64600 | |
| }, | |
| { | |
| "epoch": 18.815673032137866, | |
| "grad_norm": 0.387746661901474, | |
| "learning_rate": 0.0003744356435643564, | |
| "loss": 3.2458, | |
| "step": 64650 | |
| }, | |
| { | |
| "epoch": 18.830228225430833, | |
| "grad_norm": 0.37255313992500305, | |
| "learning_rate": 0.00037426092020966797, | |
| "loss": 3.2597, | |
| "step": 64700 | |
| }, | |
| { | |
| "epoch": 18.8447834187238, | |
| "grad_norm": 0.37015074491500854, | |
| "learning_rate": 0.00037408619685497957, | |
| "loss": 3.2722, | |
| "step": 64750 | |
| }, | |
| { | |
| "epoch": 18.859338612016767, | |
| "grad_norm": 0.3755817413330078, | |
| "learning_rate": 0.0003739114735002912, | |
| "loss": 3.2613, | |
| "step": 64800 | |
| }, | |
| { | |
| "epoch": 18.873893805309734, | |
| "grad_norm": 0.3893248438835144, | |
| "learning_rate": 0.0003737367501456028, | |
| "loss": 3.2618, | |
| "step": 64850 | |
| }, | |
| { | |
| "epoch": 18.8884489986027, | |
| "grad_norm": 0.39064499735832214, | |
| "learning_rate": 0.00037356202679091435, | |
| "loss": 3.2725, | |
| "step": 64900 | |
| }, | |
| { | |
| "epoch": 18.90300419189567, | |
| "grad_norm": 0.38426411151885986, | |
| "learning_rate": 0.00037338730343622594, | |
| "loss": 3.2659, | |
| "step": 64950 | |
| }, | |
| { | |
| "epoch": 18.917559385188635, | |
| "grad_norm": 0.3920493423938751, | |
| "learning_rate": 0.00037321258008153754, | |
| "loss": 3.2742, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.917559385188635, | |
| "eval_accuracy": 0.37444928173774933, | |
| "eval_loss": 3.527984857559204, | |
| "eval_runtime": 83.9634, | |
| "eval_samples_per_second": 198.313, | |
| "eval_steps_per_second": 12.398, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 18.932114578481603, | |
| "grad_norm": 0.38560622930526733, | |
| "learning_rate": 0.0003730378567268491, | |
| "loss": 3.2607, | |
| "step": 65050 | |
| }, | |
| { | |
| "epoch": 18.94666977177457, | |
| "grad_norm": 0.384427934885025, | |
| "learning_rate": 0.0003728631333721607, | |
| "loss": 3.2609, | |
| "step": 65100 | |
| }, | |
| { | |
| "epoch": 18.961224965067537, | |
| "grad_norm": 0.4193480312824249, | |
| "learning_rate": 0.0003726884100174723, | |
| "loss": 3.2759, | |
| "step": 65150 | |
| }, | |
| { | |
| "epoch": 18.975780158360504, | |
| "grad_norm": 0.3602101504802704, | |
| "learning_rate": 0.0003725136866627839, | |
| "loss": 3.2744, | |
| "step": 65200 | |
| }, | |
| { | |
| "epoch": 18.99033535165347, | |
| "grad_norm": 0.3758927881717682, | |
| "learning_rate": 0.00037233896330809545, | |
| "loss": 3.2744, | |
| "step": 65250 | |
| }, | |
| { | |
| "epoch": 19.00465766185375, | |
| "grad_norm": 0.42186909914016724, | |
| "learning_rate": 0.00037216423995340705, | |
| "loss": 3.2336, | |
| "step": 65300 | |
| }, | |
| { | |
| "epoch": 19.019212855146716, | |
| "grad_norm": 0.4099094271659851, | |
| "learning_rate": 0.0003719895165987187, | |
| "loss": 3.1617, | |
| "step": 65350 | |
| }, | |
| { | |
| "epoch": 19.033768048439683, | |
| "grad_norm": 0.3768123686313629, | |
| "learning_rate": 0.0003718147932440303, | |
| "loss": 3.1718, | |
| "step": 65400 | |
| }, | |
| { | |
| "epoch": 19.04832324173265, | |
| "grad_norm": 0.4209027588367462, | |
| "learning_rate": 0.00037164006988934183, | |
| "loss": 3.1765, | |
| "step": 65450 | |
| }, | |
| { | |
| "epoch": 19.062878435025617, | |
| "grad_norm": 0.36871790885925293, | |
| "learning_rate": 0.0003714653465346534, | |
| "loss": 3.1689, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 19.077433628318584, | |
| "grad_norm": 0.40158751606941223, | |
| "learning_rate": 0.000371290623179965, | |
| "loss": 3.1888, | |
| "step": 65550 | |
| }, | |
| { | |
| "epoch": 19.09198882161155, | |
| "grad_norm": 0.40643566846847534, | |
| "learning_rate": 0.00037111589982527656, | |
| "loss": 3.1794, | |
| "step": 65600 | |
| }, | |
| { | |
| "epoch": 19.106544014904518, | |
| "grad_norm": 0.3684704005718231, | |
| "learning_rate": 0.0003709411764705882, | |
| "loss": 3.1869, | |
| "step": 65650 | |
| }, | |
| { | |
| "epoch": 19.121099208197485, | |
| "grad_norm": 0.39968571066856384, | |
| "learning_rate": 0.0003707664531158998, | |
| "loss": 3.1875, | |
| "step": 65700 | |
| }, | |
| { | |
| "epoch": 19.135654401490452, | |
| "grad_norm": 0.36954888701438904, | |
| "learning_rate": 0.0003705917297612114, | |
| "loss": 3.1926, | |
| "step": 65750 | |
| }, | |
| { | |
| "epoch": 19.15020959478342, | |
| "grad_norm": 0.37278667092323303, | |
| "learning_rate": 0.000370417006406523, | |
| "loss": 3.1955, | |
| "step": 65800 | |
| }, | |
| { | |
| "epoch": 19.164764788076386, | |
| "grad_norm": 0.3847843408584595, | |
| "learning_rate": 0.00037024228305183453, | |
| "loss": 3.1886, | |
| "step": 65850 | |
| }, | |
| { | |
| "epoch": 19.179319981369353, | |
| "grad_norm": 0.43063992261886597, | |
| "learning_rate": 0.0003700675596971461, | |
| "loss": 3.1951, | |
| "step": 65900 | |
| }, | |
| { | |
| "epoch": 19.19387517466232, | |
| "grad_norm": 0.41685760021209717, | |
| "learning_rate": 0.0003698928363424578, | |
| "loss": 3.1899, | |
| "step": 65950 | |
| }, | |
| { | |
| "epoch": 19.208430367955287, | |
| "grad_norm": 0.41611719131469727, | |
| "learning_rate": 0.00036971811298776937, | |
| "loss": 3.1909, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.208430367955287, | |
| "eval_accuracy": 0.3730138032294388, | |
| "eval_loss": 3.5466461181640625, | |
| "eval_runtime": 83.9999, | |
| "eval_samples_per_second": 198.226, | |
| "eval_steps_per_second": 12.393, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 19.222985561248255, | |
| "grad_norm": 0.3811439275741577, | |
| "learning_rate": 0.0003695433896330809, | |
| "loss": 3.2009, | |
| "step": 66050 | |
| }, | |
| { | |
| "epoch": 19.23754075454122, | |
| "grad_norm": 0.3798043429851532, | |
| "learning_rate": 0.0003693686662783925, | |
| "loss": 3.2157, | |
| "step": 66100 | |
| }, | |
| { | |
| "epoch": 19.25209594783419, | |
| "grad_norm": 0.402682363986969, | |
| "learning_rate": 0.0003691939429237041, | |
| "loss": 3.1996, | |
| "step": 66150 | |
| }, | |
| { | |
| "epoch": 19.266651141127156, | |
| "grad_norm": 0.4081108272075653, | |
| "learning_rate": 0.00036901921956901575, | |
| "loss": 3.2097, | |
| "step": 66200 | |
| }, | |
| { | |
| "epoch": 19.281206334420123, | |
| "grad_norm": 0.3806810677051544, | |
| "learning_rate": 0.0003688444962143273, | |
| "loss": 3.2125, | |
| "step": 66250 | |
| }, | |
| { | |
| "epoch": 19.29576152771309, | |
| "grad_norm": 0.3911636173725128, | |
| "learning_rate": 0.0003686697728596389, | |
| "loss": 3.2091, | |
| "step": 66300 | |
| }, | |
| { | |
| "epoch": 19.310316721006053, | |
| "grad_norm": 0.39697709679603577, | |
| "learning_rate": 0.0003684950495049505, | |
| "loss": 3.2086, | |
| "step": 66350 | |
| }, | |
| { | |
| "epoch": 19.32487191429902, | |
| "grad_norm": 0.381651371717453, | |
| "learning_rate": 0.000368320326150262, | |
| "loss": 3.2228, | |
| "step": 66400 | |
| }, | |
| { | |
| "epoch": 19.339427107591987, | |
| "grad_norm": 0.37206411361694336, | |
| "learning_rate": 0.0003681456027955736, | |
| "loss": 3.2168, | |
| "step": 66450 | |
| }, | |
| { | |
| "epoch": 19.353982300884955, | |
| "grad_norm": 0.38379523158073425, | |
| "learning_rate": 0.00036797087944088526, | |
| "loss": 3.2111, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 19.36853749417792, | |
| "grad_norm": 0.41130995750427246, | |
| "learning_rate": 0.00036779615608619685, | |
| "loss": 3.2235, | |
| "step": 66550 | |
| }, | |
| { | |
| "epoch": 19.38309268747089, | |
| "grad_norm": 0.39777064323425293, | |
| "learning_rate": 0.0003676214327315084, | |
| "loss": 3.2285, | |
| "step": 66600 | |
| }, | |
| { | |
| "epoch": 19.397647880763856, | |
| "grad_norm": 0.39666748046875, | |
| "learning_rate": 0.00036744670937682, | |
| "loss": 3.2239, | |
| "step": 66650 | |
| }, | |
| { | |
| "epoch": 19.412203074056823, | |
| "grad_norm": 0.406535267829895, | |
| "learning_rate": 0.0003672719860221316, | |
| "loss": 3.2116, | |
| "step": 66700 | |
| }, | |
| { | |
| "epoch": 19.42675826734979, | |
| "grad_norm": 0.38417983055114746, | |
| "learning_rate": 0.00036709726266744323, | |
| "loss": 3.2443, | |
| "step": 66750 | |
| }, | |
| { | |
| "epoch": 19.441313460642757, | |
| "grad_norm": 0.3597206771373749, | |
| "learning_rate": 0.00036692253931275477, | |
| "loss": 3.2119, | |
| "step": 66800 | |
| }, | |
| { | |
| "epoch": 19.455868653935724, | |
| "grad_norm": 0.4211217164993286, | |
| "learning_rate": 0.00036674781595806636, | |
| "loss": 3.2221, | |
| "step": 66850 | |
| }, | |
| { | |
| "epoch": 19.47042384722869, | |
| "grad_norm": 0.3983757197856903, | |
| "learning_rate": 0.00036657309260337796, | |
| "loss": 3.2298, | |
| "step": 66900 | |
| }, | |
| { | |
| "epoch": 19.484979040521658, | |
| "grad_norm": 0.3916017413139343, | |
| "learning_rate": 0.00036639836924868955, | |
| "loss": 3.225, | |
| "step": 66950 | |
| }, | |
| { | |
| "epoch": 19.499534233814625, | |
| "grad_norm": 0.38484591245651245, | |
| "learning_rate": 0.0003662236458940011, | |
| "loss": 3.2298, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.499534233814625, | |
| "eval_accuracy": 0.3736470403510023, | |
| "eval_loss": 3.5387115478515625, | |
| "eval_runtime": 84.0784, | |
| "eval_samples_per_second": 198.041, | |
| "eval_steps_per_second": 12.381, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 19.514089427107592, | |
| "grad_norm": 0.39086875319480896, | |
| "learning_rate": 0.00036604892253931274, | |
| "loss": 3.2291, | |
| "step": 67050 | |
| }, | |
| { | |
| "epoch": 19.52864462040056, | |
| "grad_norm": 0.3944292366504669, | |
| "learning_rate": 0.00036587419918462433, | |
| "loss": 3.2332, | |
| "step": 67100 | |
| }, | |
| { | |
| "epoch": 19.543199813693526, | |
| "grad_norm": 0.3889774680137634, | |
| "learning_rate": 0.00036569947582993593, | |
| "loss": 3.2388, | |
| "step": 67150 | |
| }, | |
| { | |
| "epoch": 19.557755006986493, | |
| "grad_norm": 0.3904835879802704, | |
| "learning_rate": 0.00036552475247524747, | |
| "loss": 3.2397, | |
| "step": 67200 | |
| }, | |
| { | |
| "epoch": 19.57231020027946, | |
| "grad_norm": 0.3771398365497589, | |
| "learning_rate": 0.00036535002912055906, | |
| "loss": 3.2294, | |
| "step": 67250 | |
| }, | |
| { | |
| "epoch": 19.586865393572428, | |
| "grad_norm": 0.4034382998943329, | |
| "learning_rate": 0.00036517530576587066, | |
| "loss": 3.2425, | |
| "step": 67300 | |
| }, | |
| { | |
| "epoch": 19.601420586865395, | |
| "grad_norm": 0.3894699811935425, | |
| "learning_rate": 0.0003650005824111823, | |
| "loss": 3.2435, | |
| "step": 67350 | |
| }, | |
| { | |
| "epoch": 19.61597578015836, | |
| "grad_norm": 0.38402053713798523, | |
| "learning_rate": 0.00036482585905649385, | |
| "loss": 3.2433, | |
| "step": 67400 | |
| }, | |
| { | |
| "epoch": 19.63053097345133, | |
| "grad_norm": 0.3767937421798706, | |
| "learning_rate": 0.00036465113570180544, | |
| "loss": 3.248, | |
| "step": 67450 | |
| }, | |
| { | |
| "epoch": 19.645086166744296, | |
| "grad_norm": 0.3979004919528961, | |
| "learning_rate": 0.00036447641234711703, | |
| "loss": 3.2419, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 19.659641360037263, | |
| "grad_norm": 0.3611491918563843, | |
| "learning_rate": 0.0003643016889924286, | |
| "loss": 3.2554, | |
| "step": 67550 | |
| }, | |
| { | |
| "epoch": 19.67419655333023, | |
| "grad_norm": 0.3876889944076538, | |
| "learning_rate": 0.0003641269656377402, | |
| "loss": 3.2481, | |
| "step": 67600 | |
| }, | |
| { | |
| "epoch": 19.688751746623197, | |
| "grad_norm": 0.39058175683021545, | |
| "learning_rate": 0.0003639522422830518, | |
| "loss": 3.2411, | |
| "step": 67650 | |
| }, | |
| { | |
| "epoch": 19.70330693991616, | |
| "grad_norm": 0.35966378450393677, | |
| "learning_rate": 0.0003637775189283634, | |
| "loss": 3.2389, | |
| "step": 67700 | |
| }, | |
| { | |
| "epoch": 19.717862133209128, | |
| "grad_norm": 0.3897443115711212, | |
| "learning_rate": 0.00036360279557367495, | |
| "loss": 3.2423, | |
| "step": 67750 | |
| }, | |
| { | |
| "epoch": 19.732417326502095, | |
| "grad_norm": 0.4152631163597107, | |
| "learning_rate": 0.00036342807221898655, | |
| "loss": 3.2468, | |
| "step": 67800 | |
| }, | |
| { | |
| "epoch": 19.74697251979506, | |
| "grad_norm": 0.4065493047237396, | |
| "learning_rate": 0.00036325334886429814, | |
| "loss": 3.2336, | |
| "step": 67850 | |
| }, | |
| { | |
| "epoch": 19.76152771308803, | |
| "grad_norm": 0.39475610852241516, | |
| "learning_rate": 0.0003630786255096098, | |
| "loss": 3.2389, | |
| "step": 67900 | |
| }, | |
| { | |
| "epoch": 19.776082906380996, | |
| "grad_norm": 0.40090420842170715, | |
| "learning_rate": 0.00036290390215492133, | |
| "loss": 3.2645, | |
| "step": 67950 | |
| }, | |
| { | |
| "epoch": 19.790638099673963, | |
| "grad_norm": 0.410026878118515, | |
| "learning_rate": 0.0003627291788002329, | |
| "loss": 3.2633, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.790638099673963, | |
| "eval_accuracy": 0.3743361029426369, | |
| "eval_loss": 3.532797336578369, | |
| "eval_runtime": 84.0005, | |
| "eval_samples_per_second": 198.225, | |
| "eval_steps_per_second": 12.393, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 19.80519329296693, | |
| "grad_norm": 0.3888506293296814, | |
| "learning_rate": 0.0003625544554455445, | |
| "loss": 3.2499, | |
| "step": 68050 | |
| }, | |
| { | |
| "epoch": 19.819748486259897, | |
| "grad_norm": 0.4230787754058838, | |
| "learning_rate": 0.0003623797320908561, | |
| "loss": 3.2487, | |
| "step": 68100 | |
| }, | |
| { | |
| "epoch": 19.834303679552864, | |
| "grad_norm": 0.3753487169742584, | |
| "learning_rate": 0.00036220500873616776, | |
| "loss": 3.2553, | |
| "step": 68150 | |
| }, | |
| { | |
| "epoch": 19.84885887284583, | |
| "grad_norm": 0.396729052066803, | |
| "learning_rate": 0.0003620302853814793, | |
| "loss": 3.245, | |
| "step": 68200 | |
| }, | |
| { | |
| "epoch": 19.863414066138798, | |
| "grad_norm": 0.3830254077911377, | |
| "learning_rate": 0.0003618555620267909, | |
| "loss": 3.2519, | |
| "step": 68250 | |
| }, | |
| { | |
| "epoch": 19.877969259431765, | |
| "grad_norm": 0.3637100160121918, | |
| "learning_rate": 0.0003616808386721025, | |
| "loss": 3.245, | |
| "step": 68300 | |
| }, | |
| { | |
| "epoch": 19.892524452724732, | |
| "grad_norm": 0.36592957377433777, | |
| "learning_rate": 0.00036150611531741403, | |
| "loss": 3.2531, | |
| "step": 68350 | |
| }, | |
| { | |
| "epoch": 19.9070796460177, | |
| "grad_norm": 0.38231244683265686, | |
| "learning_rate": 0.0003613313919627256, | |
| "loss": 3.2577, | |
| "step": 68400 | |
| }, | |
| { | |
| "epoch": 19.921634839310666, | |
| "grad_norm": 0.3804560601711273, | |
| "learning_rate": 0.00036115666860803727, | |
| "loss": 3.254, | |
| "step": 68450 | |
| }, | |
| { | |
| "epoch": 19.936190032603633, | |
| "grad_norm": 0.40204477310180664, | |
| "learning_rate": 0.00036098194525334887, | |
| "loss": 3.2594, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 19.9507452258966, | |
| "grad_norm": 0.38184401392936707, | |
| "learning_rate": 0.0003608072218986604, | |
| "loss": 3.2562, | |
| "step": 68550 | |
| }, | |
| { | |
| "epoch": 19.965300419189568, | |
| "grad_norm": 0.4229033887386322, | |
| "learning_rate": 0.000360632498543972, | |
| "loss": 3.2567, | |
| "step": 68600 | |
| }, | |
| { | |
| "epoch": 19.979855612482535, | |
| "grad_norm": 0.3983888626098633, | |
| "learning_rate": 0.0003604577751892836, | |
| "loss": 3.2614, | |
| "step": 68650 | |
| }, | |
| { | |
| "epoch": 19.9944108057755, | |
| "grad_norm": 0.4159657657146454, | |
| "learning_rate": 0.00036028305183459513, | |
| "loss": 3.2674, | |
| "step": 68700 | |
| }, | |
| { | |
| "epoch": 20.00873311597578, | |
| "grad_norm": 0.41202226281166077, | |
| "learning_rate": 0.0003601083284799068, | |
| "loss": 3.1972, | |
| "step": 68750 | |
| }, | |
| { | |
| "epoch": 20.023288309268747, | |
| "grad_norm": 0.3953835964202881, | |
| "learning_rate": 0.0003599336051252184, | |
| "loss": 3.1572, | |
| "step": 68800 | |
| }, | |
| { | |
| "epoch": 20.037843502561714, | |
| "grad_norm": 0.3819943070411682, | |
| "learning_rate": 0.00035975888177052997, | |
| "loss": 3.1664, | |
| "step": 68850 | |
| }, | |
| { | |
| "epoch": 20.05239869585468, | |
| "grad_norm": 0.3581252694129944, | |
| "learning_rate": 0.0003595841584158415, | |
| "loss": 3.1664, | |
| "step": 68900 | |
| }, | |
| { | |
| "epoch": 20.066953889147648, | |
| "grad_norm": 0.39779964089393616, | |
| "learning_rate": 0.0003594094350611531, | |
| "loss": 3.1653, | |
| "step": 68950 | |
| }, | |
| { | |
| "epoch": 20.081509082440615, | |
| "grad_norm": 0.400460809469223, | |
| "learning_rate": 0.00035923471170646475, | |
| "loss": 3.1564, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.081509082440615, | |
| "eval_accuracy": 0.37373130742898325, | |
| "eval_loss": 3.5451042652130127, | |
| "eval_runtime": 83.9655, | |
| "eval_samples_per_second": 198.308, | |
| "eval_steps_per_second": 12.398, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 20.096064275733582, | |
| "grad_norm": 0.4015652537345886, | |
| "learning_rate": 0.00035905998835177635, | |
| "loss": 3.167, | |
| "step": 69050 | |
| }, | |
| { | |
| "epoch": 20.11061946902655, | |
| "grad_norm": 0.4198397994041443, | |
| "learning_rate": 0.00035888526499708794, | |
| "loss": 3.1713, | |
| "step": 69100 | |
| }, | |
| { | |
| "epoch": 20.125174662319516, | |
| "grad_norm": 0.3733539879322052, | |
| "learning_rate": 0.0003587105416423995, | |
| "loss": 3.1836, | |
| "step": 69150 | |
| }, | |
| { | |
| "epoch": 20.139729855612483, | |
| "grad_norm": 0.3869462013244629, | |
| "learning_rate": 0.0003585358182877111, | |
| "loss": 3.1739, | |
| "step": 69200 | |
| }, | |
| { | |
| "epoch": 20.15428504890545, | |
| "grad_norm": 0.44308504462242126, | |
| "learning_rate": 0.00035836109493302267, | |
| "loss": 3.1861, | |
| "step": 69250 | |
| }, | |
| { | |
| "epoch": 20.168840242198417, | |
| "grad_norm": 0.3680599331855774, | |
| "learning_rate": 0.0003581863715783343, | |
| "loss": 3.1909, | |
| "step": 69300 | |
| }, | |
| { | |
| "epoch": 20.183395435491384, | |
| "grad_norm": 0.4302012622356415, | |
| "learning_rate": 0.00035801164822364586, | |
| "loss": 3.1894, | |
| "step": 69350 | |
| }, | |
| { | |
| "epoch": 20.19795062878435, | |
| "grad_norm": 0.37029576301574707, | |
| "learning_rate": 0.00035783692486895745, | |
| "loss": 3.1803, | |
| "step": 69400 | |
| }, | |
| { | |
| "epoch": 20.21250582207732, | |
| "grad_norm": 0.40083497762680054, | |
| "learning_rate": 0.00035766220151426905, | |
| "loss": 3.1942, | |
| "step": 69450 | |
| }, | |
| { | |
| "epoch": 20.227061015370285, | |
| "grad_norm": 0.41150733828544617, | |
| "learning_rate": 0.0003574874781595806, | |
| "loss": 3.1934, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 20.241616208663253, | |
| "grad_norm": 0.38415706157684326, | |
| "learning_rate": 0.00035731275480489224, | |
| "loss": 3.1941, | |
| "step": 69550 | |
| }, | |
| { | |
| "epoch": 20.25617140195622, | |
| "grad_norm": 0.40523162484169006, | |
| "learning_rate": 0.00035713803145020383, | |
| "loss": 3.1916, | |
| "step": 69600 | |
| }, | |
| { | |
| "epoch": 20.270726595249187, | |
| "grad_norm": 0.39857086539268494, | |
| "learning_rate": 0.0003569633080955154, | |
| "loss": 3.1867, | |
| "step": 69650 | |
| }, | |
| { | |
| "epoch": 20.28528178854215, | |
| "grad_norm": 0.3949422538280487, | |
| "learning_rate": 0.00035678858474082697, | |
| "loss": 3.1969, | |
| "step": 69700 | |
| }, | |
| { | |
| "epoch": 20.299836981835117, | |
| "grad_norm": 0.3775463402271271, | |
| "learning_rate": 0.00035661386138613856, | |
| "loss": 3.1925, | |
| "step": 69750 | |
| }, | |
| { | |
| "epoch": 20.314392175128084, | |
| "grad_norm": 0.3885062038898468, | |
| "learning_rate": 0.00035643913803145015, | |
| "loss": 3.1915, | |
| "step": 69800 | |
| }, | |
| { | |
| "epoch": 20.32894736842105, | |
| "grad_norm": 0.38250482082366943, | |
| "learning_rate": 0.0003562644146767618, | |
| "loss": 3.2129, | |
| "step": 69850 | |
| }, | |
| { | |
| "epoch": 20.34350256171402, | |
| "grad_norm": 0.399667352437973, | |
| "learning_rate": 0.00035608969132207334, | |
| "loss": 3.2006, | |
| "step": 69900 | |
| }, | |
| { | |
| "epoch": 20.358057755006985, | |
| "grad_norm": 0.39125746488571167, | |
| "learning_rate": 0.00035591496796738494, | |
| "loss": 3.2065, | |
| "step": 69950 | |
| }, | |
| { | |
| "epoch": 20.372612948299953, | |
| "grad_norm": 0.40468811988830566, | |
| "learning_rate": 0.00035574024461269653, | |
| "loss": 3.204, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.372612948299953, | |
| "eval_accuracy": 0.37417767613494063, | |
| "eval_loss": 3.540419816970825, | |
| "eval_runtime": 83.9445, | |
| "eval_samples_per_second": 198.357, | |
| "eval_steps_per_second": 12.401, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 20.38716814159292, | |
| "grad_norm": 0.3835304379463196, | |
| "learning_rate": 0.0003555655212580081, | |
| "loss": 3.2195, | |
| "step": 70050 | |
| }, | |
| { | |
| "epoch": 20.401723334885887, | |
| "grad_norm": 0.3872583210468292, | |
| "learning_rate": 0.00035539079790331967, | |
| "loss": 3.2165, | |
| "step": 70100 | |
| }, | |
| { | |
| "epoch": 20.416278528178854, | |
| "grad_norm": 0.37018004059791565, | |
| "learning_rate": 0.0003552160745486313, | |
| "loss": 3.2094, | |
| "step": 70150 | |
| }, | |
| { | |
| "epoch": 20.43083372147182, | |
| "grad_norm": 0.39366522431373596, | |
| "learning_rate": 0.0003550413511939429, | |
| "loss": 3.223, | |
| "step": 70200 | |
| }, | |
| { | |
| "epoch": 20.445388914764788, | |
| "grad_norm": 0.37735605239868164, | |
| "learning_rate": 0.0003548666278392545, | |
| "loss": 3.2252, | |
| "step": 70250 | |
| }, | |
| { | |
| "epoch": 20.459944108057755, | |
| "grad_norm": 0.3914068341255188, | |
| "learning_rate": 0.00035469190448456604, | |
| "loss": 3.2041, | |
| "step": 70300 | |
| }, | |
| { | |
| "epoch": 20.474499301350722, | |
| "grad_norm": 0.4133208394050598, | |
| "learning_rate": 0.00035451718112987764, | |
| "loss": 3.2194, | |
| "step": 70350 | |
| }, | |
| { | |
| "epoch": 20.48905449464369, | |
| "grad_norm": 0.39670905470848083, | |
| "learning_rate": 0.0003543424577751893, | |
| "loss": 3.2168, | |
| "step": 70400 | |
| }, | |
| { | |
| "epoch": 20.503609687936656, | |
| "grad_norm": 0.42800143361091614, | |
| "learning_rate": 0.0003541677344205009, | |
| "loss": 3.2256, | |
| "step": 70450 | |
| }, | |
| { | |
| "epoch": 20.518164881229623, | |
| "grad_norm": 0.39882606267929077, | |
| "learning_rate": 0.0003539930110658124, | |
| "loss": 3.2342, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 20.53272007452259, | |
| "grad_norm": 0.4053570628166199, | |
| "learning_rate": 0.000353818287711124, | |
| "loss": 3.2202, | |
| "step": 70550 | |
| }, | |
| { | |
| "epoch": 20.547275267815557, | |
| "grad_norm": 0.3917529284954071, | |
| "learning_rate": 0.0003536435643564356, | |
| "loss": 3.2237, | |
| "step": 70600 | |
| }, | |
| { | |
| "epoch": 20.561830461108524, | |
| "grad_norm": 0.4030987620353699, | |
| "learning_rate": 0.00035346884100174715, | |
| "loss": 3.234, | |
| "step": 70650 | |
| }, | |
| { | |
| "epoch": 20.57638565440149, | |
| "grad_norm": 0.41473761200904846, | |
| "learning_rate": 0.0003532941176470588, | |
| "loss": 3.2287, | |
| "step": 70700 | |
| }, | |
| { | |
| "epoch": 20.59094084769446, | |
| "grad_norm": 0.3932833671569824, | |
| "learning_rate": 0.0003531193942923704, | |
| "loss": 3.2344, | |
| "step": 70750 | |
| }, | |
| { | |
| "epoch": 20.605496040987425, | |
| "grad_norm": 0.40861883759498596, | |
| "learning_rate": 0.000352944670937682, | |
| "loss": 3.2351, | |
| "step": 70800 | |
| }, | |
| { | |
| "epoch": 20.620051234280393, | |
| "grad_norm": 0.4061174988746643, | |
| "learning_rate": 0.0003527699475829935, | |
| "loss": 3.2321, | |
| "step": 70850 | |
| }, | |
| { | |
| "epoch": 20.63460642757336, | |
| "grad_norm": 0.3992917835712433, | |
| "learning_rate": 0.0003525952242283051, | |
| "loss": 3.2242, | |
| "step": 70900 | |
| }, | |
| { | |
| "epoch": 20.649161620866327, | |
| "grad_norm": 0.3720845580101013, | |
| "learning_rate": 0.00035242050087361677, | |
| "loss": 3.2308, | |
| "step": 70950 | |
| }, | |
| { | |
| "epoch": 20.663716814159294, | |
| "grad_norm": 0.4167897403240204, | |
| "learning_rate": 0.00035224577751892836, | |
| "loss": 3.2406, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.663716814159294, | |
| "eval_accuracy": 0.37416698115014807, | |
| "eval_loss": 3.534292221069336, | |
| "eval_runtime": 83.9648, | |
| "eval_samples_per_second": 198.309, | |
| "eval_steps_per_second": 12.398, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 20.678272007452257, | |
| "grad_norm": 0.4045213460922241, | |
| "learning_rate": 0.0003520710541642399, | |
| "loss": 3.2337, | |
| "step": 71050 | |
| }, | |
| { | |
| "epoch": 20.692827200745224, | |
| "grad_norm": 0.37437090277671814, | |
| "learning_rate": 0.0003518963308095515, | |
| "loss": 3.2369, | |
| "step": 71100 | |
| }, | |
| { | |
| "epoch": 20.70738239403819, | |
| "grad_norm": 0.37870410084724426, | |
| "learning_rate": 0.0003517216074548631, | |
| "loss": 3.239, | |
| "step": 71150 | |
| }, | |
| { | |
| "epoch": 20.72193758733116, | |
| "grad_norm": 0.37994977831840515, | |
| "learning_rate": 0.0003515468841001747, | |
| "loss": 3.2387, | |
| "step": 71200 | |
| }, | |
| { | |
| "epoch": 20.736492780624125, | |
| "grad_norm": 0.4048125445842743, | |
| "learning_rate": 0.00035137216074548634, | |
| "loss": 3.2448, | |
| "step": 71250 | |
| }, | |
| { | |
| "epoch": 20.751047973917093, | |
| "grad_norm": 0.37109944224357605, | |
| "learning_rate": 0.0003511974373907979, | |
| "loss": 3.2301, | |
| "step": 71300 | |
| }, | |
| { | |
| "epoch": 20.76560316721006, | |
| "grad_norm": 0.403666228055954, | |
| "learning_rate": 0.00035102271403610947, | |
| "loss": 3.2392, | |
| "step": 71350 | |
| }, | |
| { | |
| "epoch": 20.780158360503027, | |
| "grad_norm": 0.37878280878067017, | |
| "learning_rate": 0.00035084799068142106, | |
| "loss": 3.2425, | |
| "step": 71400 | |
| }, | |
| { | |
| "epoch": 20.794713553795994, | |
| "grad_norm": 0.38671040534973145, | |
| "learning_rate": 0.0003506732673267326, | |
| "loss": 3.2416, | |
| "step": 71450 | |
| }, | |
| { | |
| "epoch": 20.80926874708896, | |
| "grad_norm": 0.38587531447410583, | |
| "learning_rate": 0.00035049854397204425, | |
| "loss": 3.243, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 20.823823940381928, | |
| "grad_norm": 0.3929380476474762, | |
| "learning_rate": 0.00035032382061735585, | |
| "loss": 3.2449, | |
| "step": 71550 | |
| }, | |
| { | |
| "epoch": 20.838379133674895, | |
| "grad_norm": 0.38904839754104614, | |
| "learning_rate": 0.00035014909726266744, | |
| "loss": 3.2352, | |
| "step": 71600 | |
| }, | |
| { | |
| "epoch": 20.852934326967862, | |
| "grad_norm": 0.3951546251773834, | |
| "learning_rate": 0.000349974373907979, | |
| "loss": 3.2392, | |
| "step": 71650 | |
| }, | |
| { | |
| "epoch": 20.86748952026083, | |
| "grad_norm": 0.3785591125488281, | |
| "learning_rate": 0.0003497996505532906, | |
| "loss": 3.2265, | |
| "step": 71700 | |
| }, | |
| { | |
| "epoch": 20.882044713553796, | |
| "grad_norm": 0.39375925064086914, | |
| "learning_rate": 0.00034962492719860217, | |
| "loss": 3.2556, | |
| "step": 71750 | |
| }, | |
| { | |
| "epoch": 20.896599906846763, | |
| "grad_norm": 0.4231417775154114, | |
| "learning_rate": 0.0003494502038439138, | |
| "loss": 3.2475, | |
| "step": 71800 | |
| }, | |
| { | |
| "epoch": 20.91115510013973, | |
| "grad_norm": 0.3982008397579193, | |
| "learning_rate": 0.00034927548048922536, | |
| "loss": 3.2268, | |
| "step": 71850 | |
| }, | |
| { | |
| "epoch": 20.925710293432697, | |
| "grad_norm": 0.38532185554504395, | |
| "learning_rate": 0.00034910075713453695, | |
| "loss": 3.2377, | |
| "step": 71900 | |
| }, | |
| { | |
| "epoch": 20.940265486725664, | |
| "grad_norm": 0.3894871473312378, | |
| "learning_rate": 0.00034892603377984855, | |
| "loss": 3.2524, | |
| "step": 71950 | |
| }, | |
| { | |
| "epoch": 20.95482068001863, | |
| "grad_norm": 0.3954565227031708, | |
| "learning_rate": 0.0003487513104251601, | |
| "loss": 3.2356, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.95482068001863, | |
| "eval_accuracy": 0.37458208759286565, | |
| "eval_loss": 3.529775381088257, | |
| "eval_runtime": 84.3717, | |
| "eval_samples_per_second": 197.353, | |
| "eval_steps_per_second": 12.338, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 20.9693758733116, | |
| "grad_norm": 0.37562480568885803, | |
| "learning_rate": 0.0003485765870704717, | |
| "loss": 3.2497, | |
| "step": 72050 | |
| }, | |
| { | |
| "epoch": 20.983931066604566, | |
| "grad_norm": 0.4024716019630432, | |
| "learning_rate": 0.00034840186371578333, | |
| "loss": 3.2422, | |
| "step": 72100 | |
| }, | |
| { | |
| "epoch": 20.998486259897533, | |
| "grad_norm": 0.40725913643836975, | |
| "learning_rate": 0.0003482271403610949, | |
| "loss": 3.2575, | |
| "step": 72150 | |
| }, | |
| { | |
| "epoch": 21.01280857009781, | |
| "grad_norm": 0.39202967286109924, | |
| "learning_rate": 0.0003480524170064065, | |
| "loss": 3.1558, | |
| "step": 72200 | |
| }, | |
| { | |
| "epoch": 21.027363763390778, | |
| "grad_norm": 0.4211905598640442, | |
| "learning_rate": 0.00034787769365171806, | |
| "loss": 3.1459, | |
| "step": 72250 | |
| }, | |
| { | |
| "epoch": 21.041918956683745, | |
| "grad_norm": 0.3993864059448242, | |
| "learning_rate": 0.00034770297029702965, | |
| "loss": 3.1521, | |
| "step": 72300 | |
| }, | |
| { | |
| "epoch": 21.05647414997671, | |
| "grad_norm": 0.42238959670066833, | |
| "learning_rate": 0.0003475282469423413, | |
| "loss": 3.1397, | |
| "step": 72350 | |
| }, | |
| { | |
| "epoch": 21.07102934326968, | |
| "grad_norm": 0.3965396285057068, | |
| "learning_rate": 0.0003473535235876529, | |
| "loss": 3.1484, | |
| "step": 72400 | |
| }, | |
| { | |
| "epoch": 21.085584536562646, | |
| "grad_norm": 0.4073790907859802, | |
| "learning_rate": 0.00034717880023296444, | |
| "loss": 3.1592, | |
| "step": 72450 | |
| }, | |
| { | |
| "epoch": 21.100139729855613, | |
| "grad_norm": 0.4092983603477478, | |
| "learning_rate": 0.00034700407687827603, | |
| "loss": 3.1617, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 21.11469492314858, | |
| "grad_norm": 0.42168405652046204, | |
| "learning_rate": 0.0003468293535235876, | |
| "loss": 3.1661, | |
| "step": 72550 | |
| }, | |
| { | |
| "epoch": 21.129250116441547, | |
| "grad_norm": 0.3875204026699066, | |
| "learning_rate": 0.00034665463016889916, | |
| "loss": 3.1767, | |
| "step": 72600 | |
| }, | |
| { | |
| "epoch": 21.143805309734514, | |
| "grad_norm": 0.38974329829216003, | |
| "learning_rate": 0.0003464799068142108, | |
| "loss": 3.1684, | |
| "step": 72650 | |
| }, | |
| { | |
| "epoch": 21.15836050302748, | |
| "grad_norm": 0.39823585748672485, | |
| "learning_rate": 0.0003463051834595224, | |
| "loss": 3.1742, | |
| "step": 72700 | |
| }, | |
| { | |
| "epoch": 21.172915696320448, | |
| "grad_norm": 0.4063786566257477, | |
| "learning_rate": 0.000346130460104834, | |
| "loss": 3.1757, | |
| "step": 72750 | |
| }, | |
| { | |
| "epoch": 21.187470889613415, | |
| "grad_norm": 0.4012933075428009, | |
| "learning_rate": 0.00034595573675014554, | |
| "loss": 3.1837, | |
| "step": 72800 | |
| }, | |
| { | |
| "epoch": 21.202026082906382, | |
| "grad_norm": 0.39778319001197815, | |
| "learning_rate": 0.00034578101339545714, | |
| "loss": 3.1849, | |
| "step": 72850 | |
| }, | |
| { | |
| "epoch": 21.21658127619935, | |
| "grad_norm": 0.3936227858066559, | |
| "learning_rate": 0.0003456062900407688, | |
| "loss": 3.1868, | |
| "step": 72900 | |
| }, | |
| { | |
| "epoch": 21.231136469492316, | |
| "grad_norm": 0.4074285626411438, | |
| "learning_rate": 0.0003454315666860804, | |
| "loss": 3.1853, | |
| "step": 72950 | |
| }, | |
| { | |
| "epoch": 21.245691662785283, | |
| "grad_norm": 0.42089933156967163, | |
| "learning_rate": 0.0003452568433313919, | |
| "loss": 3.1908, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.245691662785283, | |
| "eval_accuracy": 0.37369311105472414, | |
| "eval_loss": 3.54518723487854, | |
| "eval_runtime": 84.0205, | |
| "eval_samples_per_second": 198.178, | |
| "eval_steps_per_second": 12.39, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 21.260246856078247, | |
| "grad_norm": 0.39955443143844604, | |
| "learning_rate": 0.0003450821199767035, | |
| "loss": 3.1753, | |
| "step": 73050 | |
| }, | |
| { | |
| "epoch": 21.274802049371214, | |
| "grad_norm": 0.39862433075904846, | |
| "learning_rate": 0.0003449073966220151, | |
| "loss": 3.1853, | |
| "step": 73100 | |
| }, | |
| { | |
| "epoch": 21.28935724266418, | |
| "grad_norm": 0.40115195512771606, | |
| "learning_rate": 0.0003447326732673267, | |
| "loss": 3.1964, | |
| "step": 73150 | |
| }, | |
| { | |
| "epoch": 21.303912435957148, | |
| "grad_norm": 0.37581589818000793, | |
| "learning_rate": 0.0003445579499126383, | |
| "loss": 3.192, | |
| "step": 73200 | |
| }, | |
| { | |
| "epoch": 21.318467629250115, | |
| "grad_norm": 0.3887140154838562, | |
| "learning_rate": 0.0003443832265579499, | |
| "loss": 3.193, | |
| "step": 73250 | |
| }, | |
| { | |
| "epoch": 21.333022822543082, | |
| "grad_norm": 0.3983941972255707, | |
| "learning_rate": 0.0003442085032032615, | |
| "loss": 3.1975, | |
| "step": 73300 | |
| }, | |
| { | |
| "epoch": 21.34757801583605, | |
| "grad_norm": 0.4001455008983612, | |
| "learning_rate": 0.0003440337798485731, | |
| "loss": 3.2073, | |
| "step": 73350 | |
| }, | |
| { | |
| "epoch": 21.362133209129016, | |
| "grad_norm": 0.4194295406341553, | |
| "learning_rate": 0.0003438590564938846, | |
| "loss": 3.1928, | |
| "step": 73400 | |
| }, | |
| { | |
| "epoch": 21.376688402421983, | |
| "grad_norm": 0.3944680690765381, | |
| "learning_rate": 0.0003436843331391962, | |
| "loss": 3.2163, | |
| "step": 73450 | |
| }, | |
| { | |
| "epoch": 21.39124359571495, | |
| "grad_norm": 0.4224843978881836, | |
| "learning_rate": 0.00034350960978450786, | |
| "loss": 3.2022, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 21.405798789007918, | |
| "grad_norm": 0.39305299520492554, | |
| "learning_rate": 0.00034333488642981946, | |
| "loss": 3.2058, | |
| "step": 73550 | |
| }, | |
| { | |
| "epoch": 21.420353982300885, | |
| "grad_norm": 0.3803151547908783, | |
| "learning_rate": 0.000343160163075131, | |
| "loss": 3.1921, | |
| "step": 73600 | |
| }, | |
| { | |
| "epoch": 21.43490917559385, | |
| "grad_norm": 0.39776045083999634, | |
| "learning_rate": 0.0003429854397204426, | |
| "loss": 3.2012, | |
| "step": 73650 | |
| }, | |
| { | |
| "epoch": 21.44946436888682, | |
| "grad_norm": 0.3970925211906433, | |
| "learning_rate": 0.0003428107163657542, | |
| "loss": 3.2156, | |
| "step": 73700 | |
| }, | |
| { | |
| "epoch": 21.464019562179786, | |
| "grad_norm": 0.37756410241127014, | |
| "learning_rate": 0.00034263599301106583, | |
| "loss": 3.2059, | |
| "step": 73750 | |
| }, | |
| { | |
| "epoch": 21.478574755472753, | |
| "grad_norm": 0.4272429943084717, | |
| "learning_rate": 0.0003424612696563774, | |
| "loss": 3.1966, | |
| "step": 73800 | |
| }, | |
| { | |
| "epoch": 21.49312994876572, | |
| "grad_norm": 0.39363569021224976, | |
| "learning_rate": 0.00034228654630168897, | |
| "loss": 3.2182, | |
| "step": 73850 | |
| }, | |
| { | |
| "epoch": 21.507685142058687, | |
| "grad_norm": 0.3860451281070709, | |
| "learning_rate": 0.00034211182294700056, | |
| "loss": 3.2015, | |
| "step": 73900 | |
| }, | |
| { | |
| "epoch": 21.522240335351654, | |
| "grad_norm": 0.3944922089576721, | |
| "learning_rate": 0.0003419370995923121, | |
| "loss": 3.2144, | |
| "step": 73950 | |
| }, | |
| { | |
| "epoch": 21.53679552864462, | |
| "grad_norm": 0.4219294488430023, | |
| "learning_rate": 0.0003417623762376237, | |
| "loss": 3.2197, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.53679552864462, | |
| "eval_accuracy": 0.37421034872584535, | |
| "eval_loss": 3.5366764068603516, | |
| "eval_runtime": 83.9136, | |
| "eval_samples_per_second": 198.43, | |
| "eval_steps_per_second": 12.406, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 21.551350721937588, | |
| "grad_norm": 0.3957951068878174, | |
| "learning_rate": 0.00034158765288293534, | |
| "loss": 3.2099, | |
| "step": 74050 | |
| }, | |
| { | |
| "epoch": 21.565905915230555, | |
| "grad_norm": 0.3994651436805725, | |
| "learning_rate": 0.00034141292952824694, | |
| "loss": 3.2128, | |
| "step": 74100 | |
| }, | |
| { | |
| "epoch": 21.580461108523522, | |
| "grad_norm": 0.40868860483169556, | |
| "learning_rate": 0.0003412382061735585, | |
| "loss": 3.2196, | |
| "step": 74150 | |
| }, | |
| { | |
| "epoch": 21.59501630181649, | |
| "grad_norm": 0.40321916341781616, | |
| "learning_rate": 0.0003410634828188701, | |
| "loss": 3.2282, | |
| "step": 74200 | |
| }, | |
| { | |
| "epoch": 21.609571495109456, | |
| "grad_norm": 0.3896167278289795, | |
| "learning_rate": 0.00034088875946418167, | |
| "loss": 3.2128, | |
| "step": 74250 | |
| }, | |
| { | |
| "epoch": 21.624126688402423, | |
| "grad_norm": 0.3794539272785187, | |
| "learning_rate": 0.0003407140361094933, | |
| "loss": 3.2141, | |
| "step": 74300 | |
| }, | |
| { | |
| "epoch": 21.63868188169539, | |
| "grad_norm": 0.39224207401275635, | |
| "learning_rate": 0.00034053931275480486, | |
| "loss": 3.2138, | |
| "step": 74350 | |
| }, | |
| { | |
| "epoch": 21.653237074988354, | |
| "grad_norm": 0.40603047609329224, | |
| "learning_rate": 0.00034036458940011645, | |
| "loss": 3.2155, | |
| "step": 74400 | |
| }, | |
| { | |
| "epoch": 21.66779226828132, | |
| "grad_norm": 0.41021832823753357, | |
| "learning_rate": 0.00034018986604542804, | |
| "loss": 3.2215, | |
| "step": 74450 | |
| }, | |
| { | |
| "epoch": 21.682347461574288, | |
| "grad_norm": 0.42015862464904785, | |
| "learning_rate": 0.00034001514269073964, | |
| "loss": 3.2289, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 21.696902654867255, | |
| "grad_norm": 0.41719141602516174, | |
| "learning_rate": 0.0003398404193360512, | |
| "loss": 3.2269, | |
| "step": 74550 | |
| }, | |
| { | |
| "epoch": 21.711457848160222, | |
| "grad_norm": 0.3997327387332916, | |
| "learning_rate": 0.00033966569598136283, | |
| "loss": 3.2181, | |
| "step": 74600 | |
| }, | |
| { | |
| "epoch": 21.72601304145319, | |
| "grad_norm": 0.39195841550827026, | |
| "learning_rate": 0.0003394909726266744, | |
| "loss": 3.2215, | |
| "step": 74650 | |
| }, | |
| { | |
| "epoch": 21.740568234746156, | |
| "grad_norm": 0.3989555835723877, | |
| "learning_rate": 0.000339316249271986, | |
| "loss": 3.2205, | |
| "step": 74700 | |
| }, | |
| { | |
| "epoch": 21.755123428039123, | |
| "grad_norm": 0.3987559378147125, | |
| "learning_rate": 0.00033914152591729756, | |
| "loss": 3.2335, | |
| "step": 74750 | |
| }, | |
| { | |
| "epoch": 21.76967862133209, | |
| "grad_norm": 0.40153810381889343, | |
| "learning_rate": 0.00033896680256260915, | |
| "loss": 3.233, | |
| "step": 74800 | |
| }, | |
| { | |
| "epoch": 21.784233814625058, | |
| "grad_norm": 0.4722578227519989, | |
| "learning_rate": 0.00033879207920792074, | |
| "loss": 3.2269, | |
| "step": 74850 | |
| }, | |
| { | |
| "epoch": 21.798789007918025, | |
| "grad_norm": 0.37624692916870117, | |
| "learning_rate": 0.0003386173558532324, | |
| "loss": 3.2158, | |
| "step": 74900 | |
| }, | |
| { | |
| "epoch": 21.81334420121099, | |
| "grad_norm": 0.411266952753067, | |
| "learning_rate": 0.00033844263249854393, | |
| "loss": 3.2254, | |
| "step": 74950 | |
| }, | |
| { | |
| "epoch": 21.82789939450396, | |
| "grad_norm": 0.3937988579273224, | |
| "learning_rate": 0.00033826790914385553, | |
| "loss": 3.2351, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.82789939450396, | |
| "eval_accuracy": 0.37457762155526, | |
| "eval_loss": 3.529994010925293, | |
| "eval_runtime": 84.1631, | |
| "eval_samples_per_second": 197.842, | |
| "eval_steps_per_second": 12.369, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 21.842454587796926, | |
| "grad_norm": 0.36118102073669434, | |
| "learning_rate": 0.0003380931857891671, | |
| "loss": 3.2298, | |
| "step": 75050 | |
| }, | |
| { | |
| "epoch": 21.857009781089893, | |
| "grad_norm": 0.39653143286705017, | |
| "learning_rate": 0.00033791846243447866, | |
| "loss": 3.2404, | |
| "step": 75100 | |
| }, | |
| { | |
| "epoch": 21.87156497438286, | |
| "grad_norm": 0.38492098450660706, | |
| "learning_rate": 0.0003377437390797903, | |
| "loss": 3.2305, | |
| "step": 75150 | |
| }, | |
| { | |
| "epoch": 21.886120167675827, | |
| "grad_norm": 0.3939381539821625, | |
| "learning_rate": 0.0003375690157251019, | |
| "loss": 3.2342, | |
| "step": 75200 | |
| }, | |
| { | |
| "epoch": 21.900675360968794, | |
| "grad_norm": 0.39002323150634766, | |
| "learning_rate": 0.0003373942923704135, | |
| "loss": 3.227, | |
| "step": 75250 | |
| }, | |
| { | |
| "epoch": 21.91523055426176, | |
| "grad_norm": 0.42010048031806946, | |
| "learning_rate": 0.00033721956901572504, | |
| "loss": 3.2391, | |
| "step": 75300 | |
| }, | |
| { | |
| "epoch": 21.929785747554728, | |
| "grad_norm": 0.37599340081214905, | |
| "learning_rate": 0.00033704484566103663, | |
| "loss": 3.2449, | |
| "step": 75350 | |
| }, | |
| { | |
| "epoch": 21.944340940847695, | |
| "grad_norm": 0.4207255244255066, | |
| "learning_rate": 0.00033687012230634823, | |
| "loss": 3.2339, | |
| "step": 75400 | |
| }, | |
| { | |
| "epoch": 21.958896134140662, | |
| "grad_norm": 0.4161336421966553, | |
| "learning_rate": 0.0003366953989516599, | |
| "loss": 3.2417, | |
| "step": 75450 | |
| }, | |
| { | |
| "epoch": 21.97345132743363, | |
| "grad_norm": 0.420137882232666, | |
| "learning_rate": 0.00033652067559697147, | |
| "loss": 3.2346, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 21.988006520726596, | |
| "grad_norm": 0.39471960067749023, | |
| "learning_rate": 0.000336345952242283, | |
| "loss": 3.2524, | |
| "step": 75550 | |
| }, | |
| { | |
| "epoch": 22.002328830926874, | |
| "grad_norm": 0.4067346155643463, | |
| "learning_rate": 0.0003361712288875946, | |
| "loss": 3.2131, | |
| "step": 75600 | |
| }, | |
| { | |
| "epoch": 22.01688402421984, | |
| "grad_norm": 0.3875446915626526, | |
| "learning_rate": 0.0003359965055329062, | |
| "loss": 3.1104, | |
| "step": 75650 | |
| }, | |
| { | |
| "epoch": 22.03143921751281, | |
| "grad_norm": 0.38433337211608887, | |
| "learning_rate": 0.00033582178217821785, | |
| "loss": 3.1379, | |
| "step": 75700 | |
| }, | |
| { | |
| "epoch": 22.045994410805775, | |
| "grad_norm": 0.4136708080768585, | |
| "learning_rate": 0.0003356470588235294, | |
| "loss": 3.1345, | |
| "step": 75750 | |
| }, | |
| { | |
| "epoch": 22.060549604098743, | |
| "grad_norm": 0.41874960064888, | |
| "learning_rate": 0.000335472335468841, | |
| "loss": 3.1461, | |
| "step": 75800 | |
| }, | |
| { | |
| "epoch": 22.07510479739171, | |
| "grad_norm": 0.41599613428115845, | |
| "learning_rate": 0.0003352976121141526, | |
| "loss": 3.1544, | |
| "step": 75850 | |
| }, | |
| { | |
| "epoch": 22.089659990684677, | |
| "grad_norm": 0.41146954894065857, | |
| "learning_rate": 0.0003351228887594641, | |
| "loss": 3.1632, | |
| "step": 75900 | |
| }, | |
| { | |
| "epoch": 22.104215183977644, | |
| "grad_norm": 0.44748198986053467, | |
| "learning_rate": 0.0003349481654047757, | |
| "loss": 3.1636, | |
| "step": 75950 | |
| }, | |
| { | |
| "epoch": 22.11877037727061, | |
| "grad_norm": 0.3916570246219635, | |
| "learning_rate": 0.00033477344205008736, | |
| "loss": 3.1663, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.11877037727061, | |
| "eval_accuracy": 0.37407566243384244, | |
| "eval_loss": 3.5441932678222656, | |
| "eval_runtime": 83.9972, | |
| "eval_samples_per_second": 198.233, | |
| "eval_steps_per_second": 12.393, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 22.133325570563578, | |
| "grad_norm": 0.39270997047424316, | |
| "learning_rate": 0.00033459871869539895, | |
| "loss": 3.1586, | |
| "step": 76050 | |
| }, | |
| { | |
| "epoch": 22.147880763856545, | |
| "grad_norm": 0.3783777952194214, | |
| "learning_rate": 0.0003344239953407105, | |
| "loss": 3.1641, | |
| "step": 76100 | |
| }, | |
| { | |
| "epoch": 22.162435957149512, | |
| "grad_norm": 0.4145548641681671, | |
| "learning_rate": 0.0003342492719860221, | |
| "loss": 3.1521, | |
| "step": 76150 | |
| }, | |
| { | |
| "epoch": 22.17699115044248, | |
| "grad_norm": 0.3848534822463989, | |
| "learning_rate": 0.0003340745486313337, | |
| "loss": 3.1532, | |
| "step": 76200 | |
| }, | |
| { | |
| "epoch": 22.191546343735446, | |
| "grad_norm": 0.4007597267627716, | |
| "learning_rate": 0.0003338998252766453, | |
| "loss": 3.1644, | |
| "step": 76250 | |
| }, | |
| { | |
| "epoch": 22.206101537028413, | |
| "grad_norm": 0.4170171022415161, | |
| "learning_rate": 0.00033372510192195687, | |
| "loss": 3.1616, | |
| "step": 76300 | |
| }, | |
| { | |
| "epoch": 22.22065673032138, | |
| "grad_norm": 0.40451207756996155, | |
| "learning_rate": 0.00033355037856726847, | |
| "loss": 3.1771, | |
| "step": 76350 | |
| }, | |
| { | |
| "epoch": 22.235211923614347, | |
| "grad_norm": 0.40556612610816956, | |
| "learning_rate": 0.00033337565521258006, | |
| "loss": 3.1789, | |
| "step": 76400 | |
| }, | |
| { | |
| "epoch": 22.24976711690731, | |
| "grad_norm": 0.427217572927475, | |
| "learning_rate": 0.00033320093185789165, | |
| "loss": 3.1743, | |
| "step": 76450 | |
| }, | |
| { | |
| "epoch": 22.264322310200278, | |
| "grad_norm": 0.41901645064353943, | |
| "learning_rate": 0.0003330262085032032, | |
| "loss": 3.184, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 22.278877503493245, | |
| "grad_norm": 0.43935996294021606, | |
| "learning_rate": 0.00033285148514851484, | |
| "loss": 3.1788, | |
| "step": 76550 | |
| }, | |
| { | |
| "epoch": 22.293432696786212, | |
| "grad_norm": 0.3948269486427307, | |
| "learning_rate": 0.00033267676179382644, | |
| "loss": 3.1909, | |
| "step": 76600 | |
| }, | |
| { | |
| "epoch": 22.30798789007918, | |
| "grad_norm": 0.37802931666374207, | |
| "learning_rate": 0.00033250203843913803, | |
| "loss": 3.1798, | |
| "step": 76650 | |
| }, | |
| { | |
| "epoch": 22.322543083372146, | |
| "grad_norm": 0.42178353667259216, | |
| "learning_rate": 0.00033232731508444957, | |
| "loss": 3.1898, | |
| "step": 76700 | |
| }, | |
| { | |
| "epoch": 22.337098276665113, | |
| "grad_norm": 0.4055041968822479, | |
| "learning_rate": 0.00033215259172976117, | |
| "loss": 3.1794, | |
| "step": 76750 | |
| }, | |
| { | |
| "epoch": 22.35165346995808, | |
| "grad_norm": 0.42163988947868347, | |
| "learning_rate": 0.00033197786837507276, | |
| "loss": 3.181, | |
| "step": 76800 | |
| }, | |
| { | |
| "epoch": 22.366208663251047, | |
| "grad_norm": 0.4090975522994995, | |
| "learning_rate": 0.0003318031450203844, | |
| "loss": 3.1993, | |
| "step": 76850 | |
| }, | |
| { | |
| "epoch": 22.380763856544014, | |
| "grad_norm": 0.3955000042915344, | |
| "learning_rate": 0.00033162842166569595, | |
| "loss": 3.1848, | |
| "step": 76900 | |
| }, | |
| { | |
| "epoch": 22.39531904983698, | |
| "grad_norm": 0.3993707001209259, | |
| "learning_rate": 0.00033145369831100754, | |
| "loss": 3.2093, | |
| "step": 76950 | |
| }, | |
| { | |
| "epoch": 22.40987424312995, | |
| "grad_norm": 0.4260542392730713, | |
| "learning_rate": 0.00033127897495631914, | |
| "loss": 3.1905, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.40987424312995, | |
| "eval_accuracy": 0.3742516008100452, | |
| "eval_loss": 3.542874574661255, | |
| "eval_runtime": 83.9761, | |
| "eval_samples_per_second": 198.283, | |
| "eval_steps_per_second": 12.396, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 22.424429436422916, | |
| "grad_norm": 0.3868073523044586, | |
| "learning_rate": 0.0003311042516016307, | |
| "loss": 3.1946, | |
| "step": 77050 | |
| }, | |
| { | |
| "epoch": 22.438984629715883, | |
| "grad_norm": 0.4179459810256958, | |
| "learning_rate": 0.0003309295282469423, | |
| "loss": 3.2036, | |
| "step": 77100 | |
| }, | |
| { | |
| "epoch": 22.45353982300885, | |
| "grad_norm": 0.37110626697540283, | |
| "learning_rate": 0.0003307548048922539, | |
| "loss": 3.1926, | |
| "step": 77150 | |
| }, | |
| { | |
| "epoch": 22.468095016301817, | |
| "grad_norm": 0.3966369032859802, | |
| "learning_rate": 0.0003305800815375655, | |
| "loss": 3.1912, | |
| "step": 77200 | |
| }, | |
| { | |
| "epoch": 22.482650209594784, | |
| "grad_norm": 0.39656341075897217, | |
| "learning_rate": 0.00033040535818287705, | |
| "loss": 3.1984, | |
| "step": 77250 | |
| }, | |
| { | |
| "epoch": 22.49720540288775, | |
| "grad_norm": 0.4031260013580322, | |
| "learning_rate": 0.00033023063482818865, | |
| "loss": 3.2068, | |
| "step": 77300 | |
| }, | |
| { | |
| "epoch": 22.511760596180718, | |
| "grad_norm": 0.41559213399887085, | |
| "learning_rate": 0.00033005591147350024, | |
| "loss": 3.1958, | |
| "step": 77350 | |
| }, | |
| { | |
| "epoch": 22.526315789473685, | |
| "grad_norm": 0.41590771079063416, | |
| "learning_rate": 0.0003298811881188119, | |
| "loss": 3.199, | |
| "step": 77400 | |
| }, | |
| { | |
| "epoch": 22.540870982766652, | |
| "grad_norm": 0.4002304971218109, | |
| "learning_rate": 0.00032970646476412343, | |
| "loss": 3.2019, | |
| "step": 77450 | |
| }, | |
| { | |
| "epoch": 22.55542617605962, | |
| "grad_norm": 0.41335025429725647, | |
| "learning_rate": 0.000329531741409435, | |
| "loss": 3.215, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 22.569981369352586, | |
| "grad_norm": 0.3770439624786377, | |
| "learning_rate": 0.0003293570180547466, | |
| "loss": 3.2057, | |
| "step": 77550 | |
| }, | |
| { | |
| "epoch": 22.584536562645553, | |
| "grad_norm": 0.376477986574173, | |
| "learning_rate": 0.0003291822947000582, | |
| "loss": 3.204, | |
| "step": 77600 | |
| }, | |
| { | |
| "epoch": 22.59909175593852, | |
| "grad_norm": 0.3942410349845886, | |
| "learning_rate": 0.00032900757134536975, | |
| "loss": 3.2044, | |
| "step": 77650 | |
| }, | |
| { | |
| "epoch": 22.613646949231487, | |
| "grad_norm": 0.37803229689598083, | |
| "learning_rate": 0.0003288328479906814, | |
| "loss": 3.209, | |
| "step": 77700 | |
| }, | |
| { | |
| "epoch": 22.628202142524454, | |
| "grad_norm": 0.3918171525001526, | |
| "learning_rate": 0.000328658124635993, | |
| "loss": 3.2033, | |
| "step": 77750 | |
| }, | |
| { | |
| "epoch": 22.64275733581742, | |
| "grad_norm": 0.4215238690376282, | |
| "learning_rate": 0.0003284834012813046, | |
| "loss": 3.2052, | |
| "step": 77800 | |
| }, | |
| { | |
| "epoch": 22.657312529110385, | |
| "grad_norm": 0.45053625106811523, | |
| "learning_rate": 0.00032830867792661613, | |
| "loss": 3.2097, | |
| "step": 77850 | |
| }, | |
| { | |
| "epoch": 22.671867722403352, | |
| "grad_norm": 0.4062984883785248, | |
| "learning_rate": 0.0003281339545719277, | |
| "loss": 3.2234, | |
| "step": 77900 | |
| }, | |
| { | |
| "epoch": 22.68642291569632, | |
| "grad_norm": 0.39808741211891174, | |
| "learning_rate": 0.0003279592312172394, | |
| "loss": 3.2223, | |
| "step": 77950 | |
| }, | |
| { | |
| "epoch": 22.700978108989286, | |
| "grad_norm": 0.4021468162536621, | |
| "learning_rate": 0.00032778450786255097, | |
| "loss": 3.2269, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.700978108989286, | |
| "eval_accuracy": 0.3746062982177807, | |
| "eval_loss": 3.530552864074707, | |
| "eval_runtime": 84.0492, | |
| "eval_samples_per_second": 198.11, | |
| "eval_steps_per_second": 12.386, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 22.715533302282253, | |
| "grad_norm": 0.41792571544647217, | |
| "learning_rate": 0.0003276097845078625, | |
| "loss": 3.2215, | |
| "step": 78050 | |
| }, | |
| { | |
| "epoch": 22.73008849557522, | |
| "grad_norm": 0.39145269989967346, | |
| "learning_rate": 0.0003274350611531741, | |
| "loss": 3.206, | |
| "step": 78100 | |
| }, | |
| { | |
| "epoch": 22.744643688868187, | |
| "grad_norm": 0.3972797691822052, | |
| "learning_rate": 0.0003272603377984857, | |
| "loss": 3.2152, | |
| "step": 78150 | |
| }, | |
| { | |
| "epoch": 22.759198882161154, | |
| "grad_norm": 0.4054664373397827, | |
| "learning_rate": 0.00032708561444379724, | |
| "loss": 3.2183, | |
| "step": 78200 | |
| }, | |
| { | |
| "epoch": 22.77375407545412, | |
| "grad_norm": 0.41212597489356995, | |
| "learning_rate": 0.0003269108910891089, | |
| "loss": 3.209, | |
| "step": 78250 | |
| }, | |
| { | |
| "epoch": 22.78830926874709, | |
| "grad_norm": 0.44951191544532776, | |
| "learning_rate": 0.0003267361677344205, | |
| "loss": 3.2173, | |
| "step": 78300 | |
| }, | |
| { | |
| "epoch": 22.802864462040056, | |
| "grad_norm": 0.39247971773147583, | |
| "learning_rate": 0.0003265614443797321, | |
| "loss": 3.2226, | |
| "step": 78350 | |
| }, | |
| { | |
| "epoch": 22.817419655333023, | |
| "grad_norm": 0.3904021680355072, | |
| "learning_rate": 0.0003263867210250436, | |
| "loss": 3.2266, | |
| "step": 78400 | |
| }, | |
| { | |
| "epoch": 22.83197484862599, | |
| "grad_norm": 0.39938342571258545, | |
| "learning_rate": 0.0003262119976703552, | |
| "loss": 3.2188, | |
| "step": 78450 | |
| }, | |
| { | |
| "epoch": 22.846530041918957, | |
| "grad_norm": 0.39037880301475525, | |
| "learning_rate": 0.00032603727431566686, | |
| "loss": 3.2157, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 22.861085235211924, | |
| "grad_norm": 0.38518035411834717, | |
| "learning_rate": 0.00032586255096097845, | |
| "loss": 3.2191, | |
| "step": 78550 | |
| }, | |
| { | |
| "epoch": 22.87564042850489, | |
| "grad_norm": 0.38432469964027405, | |
| "learning_rate": 0.00032568782760629005, | |
| "loss": 3.2118, | |
| "step": 78600 | |
| }, | |
| { | |
| "epoch": 22.890195621797858, | |
| "grad_norm": 0.4380696713924408, | |
| "learning_rate": 0.0003255131042516016, | |
| "loss": 3.2255, | |
| "step": 78650 | |
| }, | |
| { | |
| "epoch": 22.904750815090825, | |
| "grad_norm": 0.39154985547065735, | |
| "learning_rate": 0.0003253383808969132, | |
| "loss": 3.2268, | |
| "step": 78700 | |
| }, | |
| { | |
| "epoch": 22.919306008383792, | |
| "grad_norm": 0.42954781651496887, | |
| "learning_rate": 0.0003251636575422248, | |
| "loss": 3.2368, | |
| "step": 78750 | |
| }, | |
| { | |
| "epoch": 22.93386120167676, | |
| "grad_norm": 0.40430429577827454, | |
| "learning_rate": 0.0003249889341875364, | |
| "loss": 3.2361, | |
| "step": 78800 | |
| }, | |
| { | |
| "epoch": 22.948416394969726, | |
| "grad_norm": 0.4079325199127197, | |
| "learning_rate": 0.00032481421083284796, | |
| "loss": 3.2142, | |
| "step": 78850 | |
| }, | |
| { | |
| "epoch": 22.962971588262693, | |
| "grad_norm": 0.39752864837646484, | |
| "learning_rate": 0.00032463948747815956, | |
| "loss": 3.2161, | |
| "step": 78900 | |
| }, | |
| { | |
| "epoch": 22.97752678155566, | |
| "grad_norm": 0.4169577658176422, | |
| "learning_rate": 0.00032446476412347115, | |
| "loss": 3.2314, | |
| "step": 78950 | |
| }, | |
| { | |
| "epoch": 22.992081974848627, | |
| "grad_norm": 0.3821992576122284, | |
| "learning_rate": 0.0003242900407687827, | |
| "loss": 3.2333, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 22.992081974848627, | |
| "eval_accuracy": 0.37529630102785855, | |
| "eval_loss": 3.52482271194458, | |
| "eval_runtime": 84.176, | |
| "eval_samples_per_second": 197.812, | |
| "eval_steps_per_second": 12.367, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 23.006404285048905, | |
| "grad_norm": 0.37856900691986084, | |
| "learning_rate": 0.00032411531741409434, | |
| "loss": 3.1755, | |
| "step": 79050 | |
| }, | |
| { | |
| "epoch": 23.020959478341872, | |
| "grad_norm": 0.41566768288612366, | |
| "learning_rate": 0.00032394059405940593, | |
| "loss": 3.1229, | |
| "step": 79100 | |
| }, | |
| { | |
| "epoch": 23.03551467163484, | |
| "grad_norm": 0.3936156928539276, | |
| "learning_rate": 0.00032376587070471753, | |
| "loss": 3.1246, | |
| "step": 79150 | |
| }, | |
| { | |
| "epoch": 23.050069864927806, | |
| "grad_norm": 0.3876214027404785, | |
| "learning_rate": 0.00032359114735002907, | |
| "loss": 3.1301, | |
| "step": 79200 | |
| }, | |
| { | |
| "epoch": 23.064625058220773, | |
| "grad_norm": 0.3899514377117157, | |
| "learning_rate": 0.00032341642399534066, | |
| "loss": 3.1326, | |
| "step": 79250 | |
| }, | |
| { | |
| "epoch": 23.07918025151374, | |
| "grad_norm": 0.4174843728542328, | |
| "learning_rate": 0.00032324170064065226, | |
| "loss": 3.1451, | |
| "step": 79300 | |
| }, | |
| { | |
| "epoch": 23.093735444806708, | |
| "grad_norm": 0.41307854652404785, | |
| "learning_rate": 0.0003230669772859639, | |
| "loss": 3.1456, | |
| "step": 79350 | |
| }, | |
| { | |
| "epoch": 23.108290638099675, | |
| "grad_norm": 0.392428994178772, | |
| "learning_rate": 0.00032289225393127545, | |
| "loss": 3.1401, | |
| "step": 79400 | |
| }, | |
| { | |
| "epoch": 23.12284583139264, | |
| "grad_norm": 0.4244144856929779, | |
| "learning_rate": 0.00032271753057658704, | |
| "loss": 3.1458, | |
| "step": 79450 | |
| }, | |
| { | |
| "epoch": 23.13740102468561, | |
| "grad_norm": 0.3783201277256012, | |
| "learning_rate": 0.00032254280722189863, | |
| "loss": 3.1511, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 23.151956217978576, | |
| "grad_norm": 0.4029030203819275, | |
| "learning_rate": 0.00032236808386721023, | |
| "loss": 3.1549, | |
| "step": 79550 | |
| }, | |
| { | |
| "epoch": 23.166511411271543, | |
| "grad_norm": 0.3899061381816864, | |
| "learning_rate": 0.00032219336051252177, | |
| "loss": 3.1501, | |
| "step": 79600 | |
| }, | |
| { | |
| "epoch": 23.18106660456451, | |
| "grad_norm": 0.4384586215019226, | |
| "learning_rate": 0.0003220186371578334, | |
| "loss": 3.151, | |
| "step": 79650 | |
| }, | |
| { | |
| "epoch": 23.195621797857477, | |
| "grad_norm": 0.39860251545906067, | |
| "learning_rate": 0.000321843913803145, | |
| "loss": 3.159, | |
| "step": 79700 | |
| }, | |
| { | |
| "epoch": 23.210176991150444, | |
| "grad_norm": 0.4482727348804474, | |
| "learning_rate": 0.0003216691904484566, | |
| "loss": 3.1664, | |
| "step": 79750 | |
| }, | |
| { | |
| "epoch": 23.22473218444341, | |
| "grad_norm": 0.42729952931404114, | |
| "learning_rate": 0.00032149446709376815, | |
| "loss": 3.1789, | |
| "step": 79800 | |
| }, | |
| { | |
| "epoch": 23.239287377736375, | |
| "grad_norm": 0.4260234534740448, | |
| "learning_rate": 0.00032131974373907974, | |
| "loss": 3.1637, | |
| "step": 79850 | |
| }, | |
| { | |
| "epoch": 23.25384257102934, | |
| "grad_norm": 0.41321444511413574, | |
| "learning_rate": 0.0003211450203843914, | |
| "loss": 3.1544, | |
| "step": 79900 | |
| }, | |
| { | |
| "epoch": 23.26839776432231, | |
| "grad_norm": 0.4217228293418884, | |
| "learning_rate": 0.000320970297029703, | |
| "loss": 3.1684, | |
| "step": 79950 | |
| }, | |
| { | |
| "epoch": 23.282952957615276, | |
| "grad_norm": 0.4344976544380188, | |
| "learning_rate": 0.0003207955736750145, | |
| "loss": 3.1703, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.282952957615276, | |
| "eval_accuracy": 0.37414664892631166, | |
| "eval_loss": 3.5430173873901367, | |
| "eval_runtime": 83.9748, | |
| "eval_samples_per_second": 198.286, | |
| "eval_steps_per_second": 12.397, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 23.297508150908243, | |
| "grad_norm": 0.4261675477027893, | |
| "learning_rate": 0.0003206208503203261, | |
| "loss": 3.1711, | |
| "step": 80050 | |
| }, | |
| { | |
| "epoch": 23.31206334420121, | |
| "grad_norm": 0.4397713840007782, | |
| "learning_rate": 0.0003204461269656377, | |
| "loss": 3.1685, | |
| "step": 80100 | |
| }, | |
| { | |
| "epoch": 23.326618537494177, | |
| "grad_norm": 0.4119091331958771, | |
| "learning_rate": 0.00032027140361094925, | |
| "loss": 3.1856, | |
| "step": 80150 | |
| }, | |
| { | |
| "epoch": 23.341173730787144, | |
| "grad_norm": 0.4183413088321686, | |
| "learning_rate": 0.0003200966802562609, | |
| "loss": 3.1764, | |
| "step": 80200 | |
| }, | |
| { | |
| "epoch": 23.35572892408011, | |
| "grad_norm": 0.42540425062179565, | |
| "learning_rate": 0.0003199219569015725, | |
| "loss": 3.1813, | |
| "step": 80250 | |
| }, | |
| { | |
| "epoch": 23.370284117373078, | |
| "grad_norm": 0.3964354991912842, | |
| "learning_rate": 0.0003197472335468841, | |
| "loss": 3.1889, | |
| "step": 80300 | |
| }, | |
| { | |
| "epoch": 23.384839310666045, | |
| "grad_norm": 0.391984760761261, | |
| "learning_rate": 0.00031957251019219563, | |
| "loss": 3.1863, | |
| "step": 80350 | |
| }, | |
| { | |
| "epoch": 23.399394503959012, | |
| "grad_norm": 0.4097630977630615, | |
| "learning_rate": 0.0003193977868375072, | |
| "loss": 3.1854, | |
| "step": 80400 | |
| }, | |
| { | |
| "epoch": 23.41394969725198, | |
| "grad_norm": 0.42397865653038025, | |
| "learning_rate": 0.00031922306348281887, | |
| "loss": 3.1922, | |
| "step": 80450 | |
| }, | |
| { | |
| "epoch": 23.428504890544946, | |
| "grad_norm": 0.4164921045303345, | |
| "learning_rate": 0.00031904834012813047, | |
| "loss": 3.1707, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 23.443060083837914, | |
| "grad_norm": 0.4060786962509155, | |
| "learning_rate": 0.000318873616773442, | |
| "loss": 3.1928, | |
| "step": 80550 | |
| }, | |
| { | |
| "epoch": 23.45761527713088, | |
| "grad_norm": 0.39714428782463074, | |
| "learning_rate": 0.0003186988934187536, | |
| "loss": 3.1823, | |
| "step": 80600 | |
| }, | |
| { | |
| "epoch": 23.472170470423848, | |
| "grad_norm": 0.39186668395996094, | |
| "learning_rate": 0.0003185241700640652, | |
| "loss": 3.1849, | |
| "step": 80650 | |
| }, | |
| { | |
| "epoch": 23.486725663716815, | |
| "grad_norm": 0.3837997317314148, | |
| "learning_rate": 0.0003183494467093768, | |
| "loss": 3.1931, | |
| "step": 80700 | |
| }, | |
| { | |
| "epoch": 23.50128085700978, | |
| "grad_norm": 0.40215760469436646, | |
| "learning_rate": 0.0003181747233546884, | |
| "loss": 3.1866, | |
| "step": 80750 | |
| }, | |
| { | |
| "epoch": 23.51583605030275, | |
| "grad_norm": 0.4194908142089844, | |
| "learning_rate": 0.000318, | |
| "loss": 3.1809, | |
| "step": 80800 | |
| }, | |
| { | |
| "epoch": 23.530391243595716, | |
| "grad_norm": 0.3897694945335388, | |
| "learning_rate": 0.00031782527664531157, | |
| "loss": 3.1814, | |
| "step": 80850 | |
| }, | |
| { | |
| "epoch": 23.544946436888683, | |
| "grad_norm": 0.392191082239151, | |
| "learning_rate": 0.00031765055329062317, | |
| "loss": 3.1978, | |
| "step": 80900 | |
| }, | |
| { | |
| "epoch": 23.55950163018165, | |
| "grad_norm": 0.39261746406555176, | |
| "learning_rate": 0.0003174758299359347, | |
| "loss": 3.1958, | |
| "step": 80950 | |
| }, | |
| { | |
| "epoch": 23.574056823474617, | |
| "grad_norm": 0.43469882011413574, | |
| "learning_rate": 0.0003173011065812463, | |
| "loss": 3.2025, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.574056823474617, | |
| "eval_accuracy": 0.37450240407979585, | |
| "eval_loss": 3.5374250411987305, | |
| "eval_runtime": 84.0253, | |
| "eval_samples_per_second": 198.167, | |
| "eval_steps_per_second": 12.389, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 23.588612016767584, | |
| "grad_norm": 0.42261895537376404, | |
| "learning_rate": 0.00031712638322655795, | |
| "loss": 3.1939, | |
| "step": 81050 | |
| }, | |
| { | |
| "epoch": 23.60316721006055, | |
| "grad_norm": 0.41939258575439453, | |
| "learning_rate": 0.00031695165987186954, | |
| "loss": 3.1978, | |
| "step": 81100 | |
| }, | |
| { | |
| "epoch": 23.61772240335352, | |
| "grad_norm": 0.3738189935684204, | |
| "learning_rate": 0.0003167769365171811, | |
| "loss": 3.202, | |
| "step": 81150 | |
| }, | |
| { | |
| "epoch": 23.63227759664648, | |
| "grad_norm": 0.39011070132255554, | |
| "learning_rate": 0.0003166022131624927, | |
| "loss": 3.2019, | |
| "step": 81200 | |
| }, | |
| { | |
| "epoch": 23.64683278993945, | |
| "grad_norm": 0.40499037504196167, | |
| "learning_rate": 0.00031642748980780427, | |
| "loss": 3.2152, | |
| "step": 81250 | |
| }, | |
| { | |
| "epoch": 23.661387983232416, | |
| "grad_norm": 0.43408268690109253, | |
| "learning_rate": 0.0003162527664531159, | |
| "loss": 3.2123, | |
| "step": 81300 | |
| }, | |
| { | |
| "epoch": 23.675943176525383, | |
| "grad_norm": 0.416657418012619, | |
| "learning_rate": 0.00031607804309842746, | |
| "loss": 3.2022, | |
| "step": 81350 | |
| }, | |
| { | |
| "epoch": 23.69049836981835, | |
| "grad_norm": 0.4164252281188965, | |
| "learning_rate": 0.00031590331974373905, | |
| "loss": 3.2012, | |
| "step": 81400 | |
| }, | |
| { | |
| "epoch": 23.705053563111317, | |
| "grad_norm": 0.4167860746383667, | |
| "learning_rate": 0.00031572859638905065, | |
| "loss": 3.1983, | |
| "step": 81450 | |
| }, | |
| { | |
| "epoch": 23.719608756404284, | |
| "grad_norm": 0.39988455176353455, | |
| "learning_rate": 0.0003155538730343622, | |
| "loss": 3.2115, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 23.73416394969725, | |
| "grad_norm": 0.408404678106308, | |
| "learning_rate": 0.0003153791496796738, | |
| "loss": 3.2033, | |
| "step": 81550 | |
| }, | |
| { | |
| "epoch": 23.74871914299022, | |
| "grad_norm": 0.3888292610645294, | |
| "learning_rate": 0.00031520442632498543, | |
| "loss": 3.2079, | |
| "step": 81600 | |
| }, | |
| { | |
| "epoch": 23.763274336283185, | |
| "grad_norm": 0.4018975496292114, | |
| "learning_rate": 0.000315029702970297, | |
| "loss": 3.2085, | |
| "step": 81650 | |
| }, | |
| { | |
| "epoch": 23.777829529576152, | |
| "grad_norm": 0.41090890765190125, | |
| "learning_rate": 0.0003148549796156086, | |
| "loss": 3.2135, | |
| "step": 81700 | |
| }, | |
| { | |
| "epoch": 23.79238472286912, | |
| "grad_norm": 0.43456175923347473, | |
| "learning_rate": 0.00031468025626092016, | |
| "loss": 3.222, | |
| "step": 81750 | |
| }, | |
| { | |
| "epoch": 23.806939916162086, | |
| "grad_norm": 0.41135847568511963, | |
| "learning_rate": 0.00031450553290623175, | |
| "loss": 3.2162, | |
| "step": 81800 | |
| }, | |
| { | |
| "epoch": 23.821495109455054, | |
| "grad_norm": 0.4160410165786743, | |
| "learning_rate": 0.0003143308095515434, | |
| "loss": 3.2149, | |
| "step": 81850 | |
| }, | |
| { | |
| "epoch": 23.83605030274802, | |
| "grad_norm": 0.4027639329433441, | |
| "learning_rate": 0.000314156086196855, | |
| "loss": 3.2074, | |
| "step": 81900 | |
| }, | |
| { | |
| "epoch": 23.850605496040988, | |
| "grad_norm": 0.4043556749820709, | |
| "learning_rate": 0.00031398136284216654, | |
| "loss": 3.2096, | |
| "step": 81950 | |
| }, | |
| { | |
| "epoch": 23.865160689333955, | |
| "grad_norm": 0.43335869908332825, | |
| "learning_rate": 0.00031380663948747813, | |
| "loss": 3.2064, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.865160689333955, | |
| "eval_accuracy": 0.37490093917245026, | |
| "eval_loss": 3.528548002243042, | |
| "eval_runtime": 84.0499, | |
| "eval_samples_per_second": 198.108, | |
| "eval_steps_per_second": 12.385, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 23.879715882626922, | |
| "grad_norm": 0.4211900234222412, | |
| "learning_rate": 0.0003136319161327897, | |
| "loss": 3.2111, | |
| "step": 82050 | |
| }, | |
| { | |
| "epoch": 23.89427107591989, | |
| "grad_norm": 0.38425305485725403, | |
| "learning_rate": 0.00031345719277810127, | |
| "loss": 3.2087, | |
| "step": 82100 | |
| }, | |
| { | |
| "epoch": 23.908826269212856, | |
| "grad_norm": 0.39567556977272034, | |
| "learning_rate": 0.0003132824694234129, | |
| "loss": 3.2053, | |
| "step": 82150 | |
| }, | |
| { | |
| "epoch": 23.923381462505823, | |
| "grad_norm": 0.4136960208415985, | |
| "learning_rate": 0.0003131077460687245, | |
| "loss": 3.2191, | |
| "step": 82200 | |
| }, | |
| { | |
| "epoch": 23.93793665579879, | |
| "grad_norm": 0.38165879249572754, | |
| "learning_rate": 0.0003129330227140361, | |
| "loss": 3.2223, | |
| "step": 82250 | |
| }, | |
| { | |
| "epoch": 23.952491849091757, | |
| "grad_norm": 0.4014659821987152, | |
| "learning_rate": 0.00031275829935934764, | |
| "loss": 3.2196, | |
| "step": 82300 | |
| }, | |
| { | |
| "epoch": 23.967047042384724, | |
| "grad_norm": 0.41406312584877014, | |
| "learning_rate": 0.00031258357600465924, | |
| "loss": 3.2118, | |
| "step": 82350 | |
| }, | |
| { | |
| "epoch": 23.98160223567769, | |
| "grad_norm": 0.3938795328140259, | |
| "learning_rate": 0.00031240885264997083, | |
| "loss": 3.2166, | |
| "step": 82400 | |
| }, | |
| { | |
| "epoch": 23.99615742897066, | |
| "grad_norm": 0.39143791794776917, | |
| "learning_rate": 0.0003122341292952825, | |
| "loss": 3.2221, | |
| "step": 82450 | |
| }, | |
| { | |
| "epoch": 24.010479739170936, | |
| "grad_norm": 0.42827358841896057, | |
| "learning_rate": 0.000312059405940594, | |
| "loss": 3.1325, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 24.025034932463903, | |
| "grad_norm": 0.41095805168151855, | |
| "learning_rate": 0.0003118846825859056, | |
| "loss": 3.1149, | |
| "step": 82550 | |
| }, | |
| { | |
| "epoch": 24.03959012575687, | |
| "grad_norm": 0.4271800220012665, | |
| "learning_rate": 0.0003117099592312172, | |
| "loss": 3.1125, | |
| "step": 82600 | |
| }, | |
| { | |
| "epoch": 24.054145319049837, | |
| "grad_norm": 0.41644829511642456, | |
| "learning_rate": 0.0003115352358765288, | |
| "loss": 3.1408, | |
| "step": 82650 | |
| }, | |
| { | |
| "epoch": 24.068700512342804, | |
| "grad_norm": 0.4339255392551422, | |
| "learning_rate": 0.0003113605125218404, | |
| "loss": 3.1285, | |
| "step": 82700 | |
| }, | |
| { | |
| "epoch": 24.08325570563577, | |
| "grad_norm": 0.4222875237464905, | |
| "learning_rate": 0.000311185789167152, | |
| "loss": 3.1335, | |
| "step": 82750 | |
| }, | |
| { | |
| "epoch": 24.09781089892874, | |
| "grad_norm": 0.4374200701713562, | |
| "learning_rate": 0.0003110110658124636, | |
| "loss": 3.143, | |
| "step": 82800 | |
| }, | |
| { | |
| "epoch": 24.112366092221706, | |
| "grad_norm": 0.4502142071723938, | |
| "learning_rate": 0.0003108363424577752, | |
| "loss": 3.1441, | |
| "step": 82850 | |
| }, | |
| { | |
| "epoch": 24.126921285514673, | |
| "grad_norm": 0.40439507365226746, | |
| "learning_rate": 0.0003106616191030867, | |
| "loss": 3.1333, | |
| "step": 82900 | |
| }, | |
| { | |
| "epoch": 24.14147647880764, | |
| "grad_norm": 0.39784109592437744, | |
| "learning_rate": 0.0003104868957483983, | |
| "loss": 3.1435, | |
| "step": 82950 | |
| }, | |
| { | |
| "epoch": 24.156031672100607, | |
| "grad_norm": 0.4412744343280792, | |
| "learning_rate": 0.00031031217239370996, | |
| "loss": 3.1452, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.156031672100607, | |
| "eval_accuracy": 0.3740590910837792, | |
| "eval_loss": 3.5460166931152344, | |
| "eval_runtime": 83.9772, | |
| "eval_samples_per_second": 198.28, | |
| "eval_steps_per_second": 12.396, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 24.170586865393574, | |
| "grad_norm": 0.4179155230522156, | |
| "learning_rate": 0.00031013744903902156, | |
| "loss": 3.1462, | |
| "step": 83050 | |
| }, | |
| { | |
| "epoch": 24.18514205868654, | |
| "grad_norm": 0.40511590242385864, | |
| "learning_rate": 0.0003099627256843331, | |
| "loss": 3.1523, | |
| "step": 83100 | |
| }, | |
| { | |
| "epoch": 24.199697251979508, | |
| "grad_norm": 0.45681291818618774, | |
| "learning_rate": 0.0003097880023296447, | |
| "loss": 3.1529, | |
| "step": 83150 | |
| }, | |
| { | |
| "epoch": 24.21425244527247, | |
| "grad_norm": 0.4244830012321472, | |
| "learning_rate": 0.0003096132789749563, | |
| "loss": 3.1466, | |
| "step": 83200 | |
| }, | |
| { | |
| "epoch": 24.22880763856544, | |
| "grad_norm": 0.4050218462944031, | |
| "learning_rate": 0.00030943855562026794, | |
| "loss": 3.1526, | |
| "step": 83250 | |
| }, | |
| { | |
| "epoch": 24.243362831858406, | |
| "grad_norm": 0.44180533289909363, | |
| "learning_rate": 0.0003092638322655795, | |
| "loss": 3.1518, | |
| "step": 83300 | |
| }, | |
| { | |
| "epoch": 24.257918025151373, | |
| "grad_norm": 0.3926882743835449, | |
| "learning_rate": 0.00030908910891089107, | |
| "loss": 3.1649, | |
| "step": 83350 | |
| }, | |
| { | |
| "epoch": 24.27247321844434, | |
| "grad_norm": 0.4205596446990967, | |
| "learning_rate": 0.00030891438555620266, | |
| "loss": 3.164, | |
| "step": 83400 | |
| }, | |
| { | |
| "epoch": 24.287028411737307, | |
| "grad_norm": 0.44306880235671997, | |
| "learning_rate": 0.0003087396622015142, | |
| "loss": 3.1736, | |
| "step": 83450 | |
| }, | |
| { | |
| "epoch": 24.301583605030274, | |
| "grad_norm": 0.4040347635746002, | |
| "learning_rate": 0.0003085649388468258, | |
| "loss": 3.1581, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 24.31613879832324, | |
| "grad_norm": 0.41002345085144043, | |
| "learning_rate": 0.00030839021549213745, | |
| "loss": 3.163, | |
| "step": 83550 | |
| }, | |
| { | |
| "epoch": 24.330693991616208, | |
| "grad_norm": 0.394443541765213, | |
| "learning_rate": 0.00030821549213744904, | |
| "loss": 3.1666, | |
| "step": 83600 | |
| }, | |
| { | |
| "epoch": 24.345249184909175, | |
| "grad_norm": 0.40768301486968994, | |
| "learning_rate": 0.0003080407687827606, | |
| "loss": 3.1701, | |
| "step": 83650 | |
| }, | |
| { | |
| "epoch": 24.359804378202142, | |
| "grad_norm": 0.429211288690567, | |
| "learning_rate": 0.0003078660454280722, | |
| "loss": 3.1729, | |
| "step": 83700 | |
| }, | |
| { | |
| "epoch": 24.37435957149511, | |
| "grad_norm": 0.44395795464515686, | |
| "learning_rate": 0.00030769132207338377, | |
| "loss": 3.1671, | |
| "step": 83750 | |
| }, | |
| { | |
| "epoch": 24.388914764788076, | |
| "grad_norm": 0.429529070854187, | |
| "learning_rate": 0.00030751659871869536, | |
| "loss": 3.1664, | |
| "step": 83800 | |
| }, | |
| { | |
| "epoch": 24.403469958081043, | |
| "grad_norm": 0.41918620467185974, | |
| "learning_rate": 0.00030734187536400696, | |
| "loss": 3.189, | |
| "step": 83850 | |
| }, | |
| { | |
| "epoch": 24.41802515137401, | |
| "grad_norm": 0.38274475932121277, | |
| "learning_rate": 0.00030716715200931855, | |
| "loss": 3.1822, | |
| "step": 83900 | |
| }, | |
| { | |
| "epoch": 24.432580344666977, | |
| "grad_norm": 0.41394317150115967, | |
| "learning_rate": 0.00030699242865463015, | |
| "loss": 3.1706, | |
| "step": 83950 | |
| }, | |
| { | |
| "epoch": 24.447135537959944, | |
| "grad_norm": 0.3997444808483124, | |
| "learning_rate": 0.00030681770529994174, | |
| "loss": 3.1751, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.447135537959944, | |
| "eval_accuracy": 0.37492750034347355, | |
| "eval_loss": 3.5390408039093018, | |
| "eval_runtime": 83.9539, | |
| "eval_samples_per_second": 198.335, | |
| "eval_steps_per_second": 12.4, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 24.46169073125291, | |
| "grad_norm": 0.43127965927124023, | |
| "learning_rate": 0.0003066429819452533, | |
| "loss": 3.186, | |
| "step": 84050 | |
| }, | |
| { | |
| "epoch": 24.47624592454588, | |
| "grad_norm": 0.4451238811016083, | |
| "learning_rate": 0.00030646825859056493, | |
| "loss": 3.1813, | |
| "step": 84100 | |
| }, | |
| { | |
| "epoch": 24.490801117838846, | |
| "grad_norm": 0.423552930355072, | |
| "learning_rate": 0.0003062935352358765, | |
| "loss": 3.1791, | |
| "step": 84150 | |
| }, | |
| { | |
| "epoch": 24.505356311131813, | |
| "grad_norm": 0.4232950806617737, | |
| "learning_rate": 0.0003061188118811881, | |
| "loss": 3.1871, | |
| "step": 84200 | |
| }, | |
| { | |
| "epoch": 24.51991150442478, | |
| "grad_norm": 0.42577290534973145, | |
| "learning_rate": 0.00030594408852649966, | |
| "loss": 3.1826, | |
| "step": 84250 | |
| }, | |
| { | |
| "epoch": 24.534466697717747, | |
| "grad_norm": 0.4124676585197449, | |
| "learning_rate": 0.00030576936517181125, | |
| "loss": 3.1814, | |
| "step": 84300 | |
| }, | |
| { | |
| "epoch": 24.549021891010714, | |
| "grad_norm": 0.4032205045223236, | |
| "learning_rate": 0.00030559464181712285, | |
| "loss": 3.1912, | |
| "step": 84350 | |
| }, | |
| { | |
| "epoch": 24.56357708430368, | |
| "grad_norm": 0.4056820571422577, | |
| "learning_rate": 0.0003054199184624345, | |
| "loss": 3.1768, | |
| "step": 84400 | |
| }, | |
| { | |
| "epoch": 24.578132277596648, | |
| "grad_norm": 0.39344698190689087, | |
| "learning_rate": 0.00030524519510774604, | |
| "loss": 3.1857, | |
| "step": 84450 | |
| }, | |
| { | |
| "epoch": 24.592687470889615, | |
| "grad_norm": 0.4196877181529999, | |
| "learning_rate": 0.00030507047175305763, | |
| "loss": 3.2062, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 24.60724266418258, | |
| "grad_norm": 0.40726086497306824, | |
| "learning_rate": 0.0003048957483983692, | |
| "loss": 3.1804, | |
| "step": 84550 | |
| }, | |
| { | |
| "epoch": 24.621797857475546, | |
| "grad_norm": 0.3890549838542938, | |
| "learning_rate": 0.00030472102504368076, | |
| "loss": 3.1834, | |
| "step": 84600 | |
| }, | |
| { | |
| "epoch": 24.636353050768513, | |
| "grad_norm": 0.4186764359474182, | |
| "learning_rate": 0.0003045463016889924, | |
| "loss": 3.1876, | |
| "step": 84650 | |
| }, | |
| { | |
| "epoch": 24.65090824406148, | |
| "grad_norm": 0.4296596944332123, | |
| "learning_rate": 0.000304371578334304, | |
| "loss": 3.1821, | |
| "step": 84700 | |
| }, | |
| { | |
| "epoch": 24.665463437354447, | |
| "grad_norm": 0.3983154892921448, | |
| "learning_rate": 0.0003041968549796156, | |
| "loss": 3.1985, | |
| "step": 84750 | |
| }, | |
| { | |
| "epoch": 24.680018630647414, | |
| "grad_norm": 0.4121738374233246, | |
| "learning_rate": 0.00030402213162492714, | |
| "loss": 3.1921, | |
| "step": 84800 | |
| }, | |
| { | |
| "epoch": 24.69457382394038, | |
| "grad_norm": 0.4093579351902008, | |
| "learning_rate": 0.00030384740827023874, | |
| "loss": 3.1948, | |
| "step": 84850 | |
| }, | |
| { | |
| "epoch": 24.709129017233348, | |
| "grad_norm": 0.415157675743103, | |
| "learning_rate": 0.00030367268491555033, | |
| "loss": 3.1898, | |
| "step": 84900 | |
| }, | |
| { | |
| "epoch": 24.723684210526315, | |
| "grad_norm": 0.409600168466568, | |
| "learning_rate": 0.000303497961560862, | |
| "loss": 3.191, | |
| "step": 84950 | |
| }, | |
| { | |
| "epoch": 24.738239403819282, | |
| "grad_norm": 0.4319539964199066, | |
| "learning_rate": 0.00030332323820617357, | |
| "loss": 3.1984, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.738239403819282, | |
| "eval_accuracy": 0.3752246093715568, | |
| "eval_loss": 3.5300133228302, | |
| "eval_runtime": 83.9137, | |
| "eval_samples_per_second": 198.43, | |
| "eval_steps_per_second": 12.406, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 24.75279459711225, | |
| "grad_norm": 0.3894908130168915, | |
| "learning_rate": 0.0003031485148514851, | |
| "loss": 3.1907, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 24.767349790405216, | |
| "grad_norm": 0.3939189016819, | |
| "learning_rate": 0.0003029737914967967, | |
| "loss": 3.196, | |
| "step": 85100 | |
| }, | |
| { | |
| "epoch": 24.781904983698183, | |
| "grad_norm": 0.40757110714912415, | |
| "learning_rate": 0.0003027990681421083, | |
| "loss": 3.1973, | |
| "step": 85150 | |
| }, | |
| { | |
| "epoch": 24.79646017699115, | |
| "grad_norm": 0.4236782491207123, | |
| "learning_rate": 0.00030262434478741984, | |
| "loss": 3.2063, | |
| "step": 85200 | |
| }, | |
| { | |
| "epoch": 24.811015370284117, | |
| "grad_norm": 0.4145187437534332, | |
| "learning_rate": 0.0003024496214327315, | |
| "loss": 3.1968, | |
| "step": 85250 | |
| }, | |
| { | |
| "epoch": 24.825570563577084, | |
| "grad_norm": 0.386033296585083, | |
| "learning_rate": 0.0003022748980780431, | |
| "loss": 3.2011, | |
| "step": 85300 | |
| }, | |
| { | |
| "epoch": 24.84012575687005, | |
| "grad_norm": 0.39922523498535156, | |
| "learning_rate": 0.0003021001747233547, | |
| "loss": 3.2072, | |
| "step": 85350 | |
| }, | |
| { | |
| "epoch": 24.85468095016302, | |
| "grad_norm": 0.43985670804977417, | |
| "learning_rate": 0.0003019254513686662, | |
| "loss": 3.2061, | |
| "step": 85400 | |
| }, | |
| { | |
| "epoch": 24.869236143455986, | |
| "grad_norm": 0.4318005442619324, | |
| "learning_rate": 0.0003017507280139778, | |
| "loss": 3.2111, | |
| "step": 85450 | |
| }, | |
| { | |
| "epoch": 24.883791336748953, | |
| "grad_norm": 0.46620821952819824, | |
| "learning_rate": 0.00030157600465928946, | |
| "loss": 3.2041, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 24.89834653004192, | |
| "grad_norm": 0.40431949496269226, | |
| "learning_rate": 0.00030140128130460106, | |
| "loss": 3.1917, | |
| "step": 85550 | |
| }, | |
| { | |
| "epoch": 24.912901723334887, | |
| "grad_norm": 0.41156139969825745, | |
| "learning_rate": 0.0003012265579499126, | |
| "loss": 3.1975, | |
| "step": 85600 | |
| }, | |
| { | |
| "epoch": 24.927456916627854, | |
| "grad_norm": 0.4178667962551117, | |
| "learning_rate": 0.0003010518345952242, | |
| "loss": 3.2172, | |
| "step": 85650 | |
| }, | |
| { | |
| "epoch": 24.94201210992082, | |
| "grad_norm": 0.4230746030807495, | |
| "learning_rate": 0.0003008771112405358, | |
| "loss": 3.2161, | |
| "step": 85700 | |
| }, | |
| { | |
| "epoch": 24.956567303213788, | |
| "grad_norm": 0.4118896424770355, | |
| "learning_rate": 0.0003007023878858473, | |
| "loss": 3.2074, | |
| "step": 85750 | |
| }, | |
| { | |
| "epoch": 24.971122496506755, | |
| "grad_norm": 0.3989180624485016, | |
| "learning_rate": 0.000300527664531159, | |
| "loss": 3.2089, | |
| "step": 85800 | |
| }, | |
| { | |
| "epoch": 24.985677689799722, | |
| "grad_norm": 0.44706228375434875, | |
| "learning_rate": 0.00030035294117647057, | |
| "loss": 3.2076, | |
| "step": 85850 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.2503042221069336, | |
| "learning_rate": 0.00030017821782178216, | |
| "loss": 3.2086, | |
| "step": 85900 | |
| }, | |
| { | |
| "epoch": 25.014555193292967, | |
| "grad_norm": 0.4144589602947235, | |
| "learning_rate": 0.00030000349446709376, | |
| "loss": 3.1099, | |
| "step": 85950 | |
| }, | |
| { | |
| "epoch": 25.029110386585934, | |
| "grad_norm": 0.4288211464881897, | |
| "learning_rate": 0.00029982877111240535, | |
| "loss": 3.1106, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.029110386585934, | |
| "eval_accuracy": 0.3743840540832453, | |
| "eval_loss": 3.547487258911133, | |
| "eval_runtime": 83.9737, | |
| "eval_samples_per_second": 198.288, | |
| "eval_steps_per_second": 12.397, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 25.0436655798789, | |
| "grad_norm": 0.42030006647109985, | |
| "learning_rate": 0.00029965404775771694, | |
| "loss": 3.1182, | |
| "step": 86050 | |
| }, | |
| { | |
| "epoch": 25.05822077317187, | |
| "grad_norm": 0.43396180868148804, | |
| "learning_rate": 0.0002994793244030285, | |
| "loss": 3.1163, | |
| "step": 86100 | |
| }, | |
| { | |
| "epoch": 25.072775966464835, | |
| "grad_norm": 0.43299299478530884, | |
| "learning_rate": 0.00029930460104834013, | |
| "loss": 3.1118, | |
| "step": 86150 | |
| }, | |
| { | |
| "epoch": 25.087331159757802, | |
| "grad_norm": 0.41128793358802795, | |
| "learning_rate": 0.0002991298776936517, | |
| "loss": 3.1155, | |
| "step": 86200 | |
| }, | |
| { | |
| "epoch": 25.10188635305077, | |
| "grad_norm": 0.40414831042289734, | |
| "learning_rate": 0.00029895515433896327, | |
| "loss": 3.1305, | |
| "step": 86250 | |
| }, | |
| { | |
| "epoch": 25.116441546343736, | |
| "grad_norm": 0.4200949966907501, | |
| "learning_rate": 0.00029878043098427486, | |
| "loss": 3.1333, | |
| "step": 86300 | |
| }, | |
| { | |
| "epoch": 25.130996739636704, | |
| "grad_norm": 0.4391709864139557, | |
| "learning_rate": 0.00029860570762958646, | |
| "loss": 3.1424, | |
| "step": 86350 | |
| }, | |
| { | |
| "epoch": 25.14555193292967, | |
| "grad_norm": 0.4387558698654175, | |
| "learning_rate": 0.00029843098427489805, | |
| "loss": 3.1431, | |
| "step": 86400 | |
| }, | |
| { | |
| "epoch": 25.160107126222638, | |
| "grad_norm": 0.44652825593948364, | |
| "learning_rate": 0.00029825626092020964, | |
| "loss": 3.1351, | |
| "step": 86450 | |
| }, | |
| { | |
| "epoch": 25.174662319515605, | |
| "grad_norm": 0.4290112257003784, | |
| "learning_rate": 0.00029808153756552124, | |
| "loss": 3.1346, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 25.189217512808572, | |
| "grad_norm": 0.4343327283859253, | |
| "learning_rate": 0.00029790681421083283, | |
| "loss": 3.1384, | |
| "step": 86550 | |
| }, | |
| { | |
| "epoch": 25.203772706101535, | |
| "grad_norm": 0.4077424108982086, | |
| "learning_rate": 0.00029773209085614443, | |
| "loss": 3.1497, | |
| "step": 86600 | |
| }, | |
| { | |
| "epoch": 25.218327899394502, | |
| "grad_norm": 0.42805057764053345, | |
| "learning_rate": 0.00029755736750145597, | |
| "loss": 3.1373, | |
| "step": 86650 | |
| }, | |
| { | |
| "epoch": 25.23288309268747, | |
| "grad_norm": 0.4265683889389038, | |
| "learning_rate": 0.0002973826441467676, | |
| "loss": 3.1554, | |
| "step": 86700 | |
| }, | |
| { | |
| "epoch": 25.247438285980436, | |
| "grad_norm": 0.4488525092601776, | |
| "learning_rate": 0.00029720792079207916, | |
| "loss": 3.1394, | |
| "step": 86750 | |
| }, | |
| { | |
| "epoch": 25.261993479273404, | |
| "grad_norm": 0.42061764001846313, | |
| "learning_rate": 0.00029703319743739075, | |
| "loss": 3.1396, | |
| "step": 86800 | |
| }, | |
| { | |
| "epoch": 25.27654867256637, | |
| "grad_norm": 0.44694581627845764, | |
| "learning_rate": 0.00029685847408270234, | |
| "loss": 3.1493, | |
| "step": 86850 | |
| }, | |
| { | |
| "epoch": 25.291103865859338, | |
| "grad_norm": 0.43014317750930786, | |
| "learning_rate": 0.00029668375072801394, | |
| "loss": 3.1558, | |
| "step": 86900 | |
| }, | |
| { | |
| "epoch": 25.305659059152305, | |
| "grad_norm": 0.42524731159210205, | |
| "learning_rate": 0.00029650902737332553, | |
| "loss": 3.1676, | |
| "step": 86950 | |
| }, | |
| { | |
| "epoch": 25.320214252445272, | |
| "grad_norm": 0.4396873712539673, | |
| "learning_rate": 0.00029633430401863713, | |
| "loss": 3.148, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.320214252445272, | |
| "eval_accuracy": 0.3741564036926609, | |
| "eval_loss": 3.5479302406311035, | |
| "eval_runtime": 83.984, | |
| "eval_samples_per_second": 198.264, | |
| "eval_steps_per_second": 12.395, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 25.33476944573824, | |
| "grad_norm": 0.4452574551105499, | |
| "learning_rate": 0.0002961595806639487, | |
| "loss": 3.1617, | |
| "step": 87050 | |
| }, | |
| { | |
| "epoch": 25.349324639031206, | |
| "grad_norm": 0.4495620131492615, | |
| "learning_rate": 0.0002959848573092603, | |
| "loss": 3.1732, | |
| "step": 87100 | |
| }, | |
| { | |
| "epoch": 25.363879832324173, | |
| "grad_norm": 0.4201694130897522, | |
| "learning_rate": 0.0002958101339545719, | |
| "loss": 3.1749, | |
| "step": 87150 | |
| }, | |
| { | |
| "epoch": 25.37843502561714, | |
| "grad_norm": 0.41929733753204346, | |
| "learning_rate": 0.0002956354105998835, | |
| "loss": 3.174, | |
| "step": 87200 | |
| }, | |
| { | |
| "epoch": 25.392990218910107, | |
| "grad_norm": 0.4319401979446411, | |
| "learning_rate": 0.0002954606872451951, | |
| "loss": 3.1588, | |
| "step": 87250 | |
| }, | |
| { | |
| "epoch": 25.407545412203074, | |
| "grad_norm": 0.4432377517223358, | |
| "learning_rate": 0.0002952859638905067, | |
| "loss": 3.1716, | |
| "step": 87300 | |
| }, | |
| { | |
| "epoch": 25.42210060549604, | |
| "grad_norm": 0.44425782561302185, | |
| "learning_rate": 0.00029511124053581823, | |
| "loss": 3.1619, | |
| "step": 87350 | |
| }, | |
| { | |
| "epoch": 25.43665579878901, | |
| "grad_norm": 0.39818328619003296, | |
| "learning_rate": 0.0002949365171811299, | |
| "loss": 3.1617, | |
| "step": 87400 | |
| }, | |
| { | |
| "epoch": 25.451210992081975, | |
| "grad_norm": 0.42208805680274963, | |
| "learning_rate": 0.0002947617938264414, | |
| "loss": 3.1633, | |
| "step": 87450 | |
| }, | |
| { | |
| "epoch": 25.465766185374942, | |
| "grad_norm": 0.4214657247066498, | |
| "learning_rate": 0.000294587070471753, | |
| "loss": 3.1758, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 25.48032137866791, | |
| "grad_norm": 0.44156280159950256, | |
| "learning_rate": 0.0002944123471170646, | |
| "loss": 3.1751, | |
| "step": 87550 | |
| }, | |
| { | |
| "epoch": 25.494876571960877, | |
| "grad_norm": 0.41614091396331787, | |
| "learning_rate": 0.0002942376237623762, | |
| "loss": 3.1861, | |
| "step": 87600 | |
| }, | |
| { | |
| "epoch": 25.509431765253844, | |
| "grad_norm": 0.4198455810546875, | |
| "learning_rate": 0.0002940629004076878, | |
| "loss": 3.1628, | |
| "step": 87650 | |
| }, | |
| { | |
| "epoch": 25.52398695854681, | |
| "grad_norm": 0.4022308588027954, | |
| "learning_rate": 0.0002938881770529994, | |
| "loss": 3.1719, | |
| "step": 87700 | |
| }, | |
| { | |
| "epoch": 25.538542151839778, | |
| "grad_norm": 0.42161357402801514, | |
| "learning_rate": 0.000293713453698311, | |
| "loss": 3.1809, | |
| "step": 87750 | |
| }, | |
| { | |
| "epoch": 25.553097345132745, | |
| "grad_norm": 0.413891077041626, | |
| "learning_rate": 0.0002935387303436226, | |
| "loss": 3.172, | |
| "step": 87800 | |
| }, | |
| { | |
| "epoch": 25.567652538425712, | |
| "grad_norm": 0.43525949120521545, | |
| "learning_rate": 0.0002933640069889342, | |
| "loss": 3.1788, | |
| "step": 87850 | |
| }, | |
| { | |
| "epoch": 25.58220773171868, | |
| "grad_norm": 0.4284988045692444, | |
| "learning_rate": 0.0002931892836342457, | |
| "loss": 3.1818, | |
| "step": 87900 | |
| }, | |
| { | |
| "epoch": 25.596762925011646, | |
| "grad_norm": 0.42671579122543335, | |
| "learning_rate": 0.00029301456027955736, | |
| "loss": 3.1708, | |
| "step": 87950 | |
| }, | |
| { | |
| "epoch": 25.61131811830461, | |
| "grad_norm": 0.4243858754634857, | |
| "learning_rate": 0.0002928398369248689, | |
| "loss": 3.1717, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.61131811830461, | |
| "eval_accuracy": 0.3751275318172859, | |
| "eval_loss": 3.5358293056488037, | |
| "eval_runtime": 84.0286, | |
| "eval_samples_per_second": 198.159, | |
| "eval_steps_per_second": 12.389, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 25.625873311597577, | |
| "grad_norm": 0.4295552372932434, | |
| "learning_rate": 0.0002926651135701805, | |
| "loss": 3.1831, | |
| "step": 88050 | |
| }, | |
| { | |
| "epoch": 25.640428504890544, | |
| "grad_norm": 0.4241816997528076, | |
| "learning_rate": 0.00029249039021549215, | |
| "loss": 3.1833, | |
| "step": 88100 | |
| }, | |
| { | |
| "epoch": 25.65498369818351, | |
| "grad_norm": 0.39900752902030945, | |
| "learning_rate": 0.0002923156668608037, | |
| "loss": 3.1811, | |
| "step": 88150 | |
| }, | |
| { | |
| "epoch": 25.669538891476478, | |
| "grad_norm": 0.46498894691467285, | |
| "learning_rate": 0.0002921409435061153, | |
| "loss": 3.1845, | |
| "step": 88200 | |
| }, | |
| { | |
| "epoch": 25.684094084769445, | |
| "grad_norm": 0.42536911368370056, | |
| "learning_rate": 0.0002919662201514269, | |
| "loss": 3.1876, | |
| "step": 88250 | |
| }, | |
| { | |
| "epoch": 25.698649278062412, | |
| "grad_norm": 0.4613916873931885, | |
| "learning_rate": 0.00029179149679673847, | |
| "loss": 3.1901, | |
| "step": 88300 | |
| }, | |
| { | |
| "epoch": 25.71320447135538, | |
| "grad_norm": 0.4010865092277527, | |
| "learning_rate": 0.00029161677344205007, | |
| "loss": 3.1902, | |
| "step": 88350 | |
| }, | |
| { | |
| "epoch": 25.727759664648346, | |
| "grad_norm": 0.4036407768726349, | |
| "learning_rate": 0.00029144205008736166, | |
| "loss": 3.1863, | |
| "step": 88400 | |
| }, | |
| { | |
| "epoch": 25.742314857941313, | |
| "grad_norm": 0.40621519088745117, | |
| "learning_rate": 0.00029126732673267325, | |
| "loss": 3.1794, | |
| "step": 88450 | |
| }, | |
| { | |
| "epoch": 25.75687005123428, | |
| "grad_norm": 0.44316810369491577, | |
| "learning_rate": 0.00029109260337798485, | |
| "loss": 3.1795, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 25.771425244527247, | |
| "grad_norm": 0.4304526448249817, | |
| "learning_rate": 0.00029091788002329644, | |
| "loss": 3.189, | |
| "step": 88550 | |
| }, | |
| { | |
| "epoch": 25.785980437820214, | |
| "grad_norm": 0.41854262351989746, | |
| "learning_rate": 0.000290743156668608, | |
| "loss": 3.1816, | |
| "step": 88600 | |
| }, | |
| { | |
| "epoch": 25.80053563111318, | |
| "grad_norm": 0.4399898946285248, | |
| "learning_rate": 0.00029056843331391963, | |
| "loss": 3.1993, | |
| "step": 88650 | |
| }, | |
| { | |
| "epoch": 25.81509082440615, | |
| "grad_norm": 0.4172545373439789, | |
| "learning_rate": 0.00029039370995923117, | |
| "loss": 3.1904, | |
| "step": 88700 | |
| }, | |
| { | |
| "epoch": 25.829646017699115, | |
| "grad_norm": 0.41036295890808105, | |
| "learning_rate": 0.00029021898660454277, | |
| "loss": 3.1892, | |
| "step": 88750 | |
| }, | |
| { | |
| "epoch": 25.844201210992082, | |
| "grad_norm": 0.40741440653800964, | |
| "learning_rate": 0.00029004426324985436, | |
| "loss": 3.1989, | |
| "step": 88800 | |
| }, | |
| { | |
| "epoch": 25.85875640428505, | |
| "grad_norm": 0.3991144001483917, | |
| "learning_rate": 0.00028986953989516595, | |
| "loss": 3.1933, | |
| "step": 88850 | |
| }, | |
| { | |
| "epoch": 25.873311597578017, | |
| "grad_norm": 0.41434815526008606, | |
| "learning_rate": 0.00028969481654047755, | |
| "loss": 3.1993, | |
| "step": 88900 | |
| }, | |
| { | |
| "epoch": 25.887866790870984, | |
| "grad_norm": 0.4150712788105011, | |
| "learning_rate": 0.00028952009318578914, | |
| "loss": 3.1801, | |
| "step": 88950 | |
| }, | |
| { | |
| "epoch": 25.90242198416395, | |
| "grad_norm": 0.4055598974227905, | |
| "learning_rate": 0.00028934536983110074, | |
| "loss": 3.2088, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.90242198416395, | |
| "eval_accuracy": 0.3754144159697983, | |
| "eval_loss": 3.529819965362549, | |
| "eval_runtime": 84.0321, | |
| "eval_samples_per_second": 198.151, | |
| "eval_steps_per_second": 12.388, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 25.916977177456918, | |
| "grad_norm": 0.45721691846847534, | |
| "learning_rate": 0.00028917064647641233, | |
| "loss": 3.1996, | |
| "step": 89050 | |
| }, | |
| { | |
| "epoch": 25.931532370749885, | |
| "grad_norm": 0.4371204376220703, | |
| "learning_rate": 0.0002889959231217239, | |
| "loss": 3.1892, | |
| "step": 89100 | |
| }, | |
| { | |
| "epoch": 25.946087564042852, | |
| "grad_norm": 0.3914988934993744, | |
| "learning_rate": 0.0002888211997670355, | |
| "loss": 3.1952, | |
| "step": 89150 | |
| }, | |
| { | |
| "epoch": 25.96064275733582, | |
| "grad_norm": 0.4025917649269104, | |
| "learning_rate": 0.0002886464764123471, | |
| "loss": 3.1998, | |
| "step": 89200 | |
| }, | |
| { | |
| "epoch": 25.975197950628786, | |
| "grad_norm": 0.40672603249549866, | |
| "learning_rate": 0.0002884717530576587, | |
| "loss": 3.2075, | |
| "step": 89250 | |
| }, | |
| { | |
| "epoch": 25.989753143921753, | |
| "grad_norm": 0.40492740273475647, | |
| "learning_rate": 0.00028829702970297025, | |
| "loss": 3.194, | |
| "step": 89300 | |
| }, | |
| { | |
| "epoch": 26.00407545412203, | |
| "grad_norm": 0.44782555103302, | |
| "learning_rate": 0.0002881223063482819, | |
| "loss": 3.164, | |
| "step": 89350 | |
| }, | |
| { | |
| "epoch": 26.018630647414998, | |
| "grad_norm": 0.4255543351173401, | |
| "learning_rate": 0.00028794758299359344, | |
| "loss": 3.1014, | |
| "step": 89400 | |
| }, | |
| { | |
| "epoch": 26.033185840707965, | |
| "grad_norm": 0.4455724358558655, | |
| "learning_rate": 0.00028777285963890503, | |
| "loss": 3.1036, | |
| "step": 89450 | |
| }, | |
| { | |
| "epoch": 26.047741034000932, | |
| "grad_norm": 0.4278687536716461, | |
| "learning_rate": 0.0002875981362842166, | |
| "loss": 3.1032, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 26.0622962272939, | |
| "grad_norm": 0.4400017559528351, | |
| "learning_rate": 0.0002874234129295282, | |
| "loss": 3.1085, | |
| "step": 89550 | |
| }, | |
| { | |
| "epoch": 26.076851420586866, | |
| "grad_norm": 0.43564850091934204, | |
| "learning_rate": 0.0002872486895748398, | |
| "loss": 3.1034, | |
| "step": 89600 | |
| }, | |
| { | |
| "epoch": 26.091406613879833, | |
| "grad_norm": 0.4188985228538513, | |
| "learning_rate": 0.0002870739662201514, | |
| "loss": 3.1125, | |
| "step": 89650 | |
| }, | |
| { | |
| "epoch": 26.1059618071728, | |
| "grad_norm": 0.41715455055236816, | |
| "learning_rate": 0.000286899242865463, | |
| "loss": 3.1201, | |
| "step": 89700 | |
| }, | |
| { | |
| "epoch": 26.120517000465767, | |
| "grad_norm": 0.4308072626590729, | |
| "learning_rate": 0.0002867245195107746, | |
| "loss": 3.1215, | |
| "step": 89750 | |
| }, | |
| { | |
| "epoch": 26.135072193758734, | |
| "grad_norm": 0.39321988821029663, | |
| "learning_rate": 0.0002865497961560862, | |
| "loss": 3.1211, | |
| "step": 89800 | |
| }, | |
| { | |
| "epoch": 26.1496273870517, | |
| "grad_norm": 0.42121532559394836, | |
| "learning_rate": 0.00028637507280139773, | |
| "loss": 3.119, | |
| "step": 89850 | |
| }, | |
| { | |
| "epoch": 26.16418258034467, | |
| "grad_norm": 0.41735926270484924, | |
| "learning_rate": 0.0002862003494467094, | |
| "loss": 3.1369, | |
| "step": 89900 | |
| }, | |
| { | |
| "epoch": 26.178737773637636, | |
| "grad_norm": 0.41900190711021423, | |
| "learning_rate": 0.0002860256260920209, | |
| "loss": 3.1424, | |
| "step": 89950 | |
| }, | |
| { | |
| "epoch": 26.1932929669306, | |
| "grad_norm": 0.4272933602333069, | |
| "learning_rate": 0.0002858509027373325, | |
| "loss": 3.1331, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.1932929669306, | |
| "eval_accuracy": 0.37478376444895384, | |
| "eval_loss": 3.546154022216797, | |
| "eval_runtime": 84.0999, | |
| "eval_samples_per_second": 197.991, | |
| "eval_steps_per_second": 12.378, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 26.207848160223566, | |
| "grad_norm": 0.47644901275634766, | |
| "learning_rate": 0.0002856761793826441, | |
| "loss": 3.1297, | |
| "step": 90050 | |
| }, | |
| { | |
| "epoch": 26.222403353516533, | |
| "grad_norm": 0.47057005763053894, | |
| "learning_rate": 0.0002855014560279557, | |
| "loss": 3.1431, | |
| "step": 90100 | |
| }, | |
| { | |
| "epoch": 26.2369585468095, | |
| "grad_norm": 0.4163896441459656, | |
| "learning_rate": 0.0002853267326732673, | |
| "loss": 3.132, | |
| "step": 90150 | |
| }, | |
| { | |
| "epoch": 26.251513740102467, | |
| "grad_norm": 0.4251641631126404, | |
| "learning_rate": 0.0002851520093185789, | |
| "loss": 3.1539, | |
| "step": 90200 | |
| }, | |
| { | |
| "epoch": 26.266068933395434, | |
| "grad_norm": 0.43715810775756836, | |
| "learning_rate": 0.0002849772859638905, | |
| "loss": 3.1498, | |
| "step": 90250 | |
| }, | |
| { | |
| "epoch": 26.2806241266884, | |
| "grad_norm": 0.5078908801078796, | |
| "learning_rate": 0.0002848025626092021, | |
| "loss": 3.1459, | |
| "step": 90300 | |
| }, | |
| { | |
| "epoch": 26.29517931998137, | |
| "grad_norm": 0.43873900175094604, | |
| "learning_rate": 0.0002846278392545137, | |
| "loss": 3.1451, | |
| "step": 90350 | |
| }, | |
| { | |
| "epoch": 26.309734513274336, | |
| "grad_norm": 0.44590896368026733, | |
| "learning_rate": 0.00028445311589982527, | |
| "loss": 3.1541, | |
| "step": 90400 | |
| }, | |
| { | |
| "epoch": 26.324289706567303, | |
| "grad_norm": 0.4568062722682953, | |
| "learning_rate": 0.00028427839254513686, | |
| "loss": 3.1647, | |
| "step": 90450 | |
| }, | |
| { | |
| "epoch": 26.33884489986027, | |
| "grad_norm": 0.42894262075424194, | |
| "learning_rate": 0.00028410366919044846, | |
| "loss": 3.1524, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 26.353400093153237, | |
| "grad_norm": 0.43518009781837463, | |
| "learning_rate": 0.00028392894583576, | |
| "loss": 3.1492, | |
| "step": 90550 | |
| }, | |
| { | |
| "epoch": 26.367955286446204, | |
| "grad_norm": 0.4703112542629242, | |
| "learning_rate": 0.00028375422248107165, | |
| "loss": 3.1569, | |
| "step": 90600 | |
| }, | |
| { | |
| "epoch": 26.38251047973917, | |
| "grad_norm": 0.41856303811073303, | |
| "learning_rate": 0.0002835794991263832, | |
| "loss": 3.145, | |
| "step": 90650 | |
| }, | |
| { | |
| "epoch": 26.397065673032138, | |
| "grad_norm": 0.4183586835861206, | |
| "learning_rate": 0.0002834047757716948, | |
| "loss": 3.1471, | |
| "step": 90700 | |
| }, | |
| { | |
| "epoch": 26.411620866325105, | |
| "grad_norm": 0.4224638342857361, | |
| "learning_rate": 0.0002832300524170064, | |
| "loss": 3.158, | |
| "step": 90750 | |
| }, | |
| { | |
| "epoch": 26.426176059618072, | |
| "grad_norm": 0.3950006365776062, | |
| "learning_rate": 0.00028305532906231797, | |
| "loss": 3.1641, | |
| "step": 90800 | |
| }, | |
| { | |
| "epoch": 26.44073125291104, | |
| "grad_norm": 0.4179461598396301, | |
| "learning_rate": 0.00028288060570762956, | |
| "loss": 3.156, | |
| "step": 90850 | |
| }, | |
| { | |
| "epoch": 26.455286446204006, | |
| "grad_norm": 0.43129533529281616, | |
| "learning_rate": 0.00028270588235294116, | |
| "loss": 3.1596, | |
| "step": 90900 | |
| }, | |
| { | |
| "epoch": 26.469841639496973, | |
| "grad_norm": 0.41973650455474854, | |
| "learning_rate": 0.00028253115899825275, | |
| "loss": 3.1497, | |
| "step": 90950 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "grad_norm": 0.44479885697364807, | |
| "learning_rate": 0.0002823564356435643, | |
| "loss": 3.162, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.48439683278994, | |
| "eval_accuracy": 0.37501823142325214, | |
| "eval_loss": 3.539034605026245, | |
| "eval_runtime": 84.0311, | |
| "eval_samples_per_second": 198.153, | |
| "eval_steps_per_second": 12.388, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 26.498952026082907, | |
| "grad_norm": 0.4507576525211334, | |
| "learning_rate": 0.00028218171228887594, | |
| "loss": 3.1719, | |
| "step": 91050 | |
| }, | |
| { | |
| "epoch": 26.513507219375875, | |
| "grad_norm": 0.42544856667518616, | |
| "learning_rate": 0.0002820069889341875, | |
| "loss": 3.1659, | |
| "step": 91100 | |
| }, | |
| { | |
| "epoch": 26.52806241266884, | |
| "grad_norm": 0.4658428132534027, | |
| "learning_rate": 0.00028183226557949913, | |
| "loss": 3.1512, | |
| "step": 91150 | |
| }, | |
| { | |
| "epoch": 26.54261760596181, | |
| "grad_norm": 0.4101416766643524, | |
| "learning_rate": 0.00028165754222481067, | |
| "loss": 3.1694, | |
| "step": 91200 | |
| }, | |
| { | |
| "epoch": 26.557172799254776, | |
| "grad_norm": 0.42220577597618103, | |
| "learning_rate": 0.00028148281887012226, | |
| "loss": 3.1672, | |
| "step": 91250 | |
| }, | |
| { | |
| "epoch": 26.571727992547743, | |
| "grad_norm": 0.4594450891017914, | |
| "learning_rate": 0.0002813080955154339, | |
| "loss": 3.1634, | |
| "step": 91300 | |
| }, | |
| { | |
| "epoch": 26.586283185840706, | |
| "grad_norm": 0.41875872015953064, | |
| "learning_rate": 0.00028113337216074545, | |
| "loss": 3.1621, | |
| "step": 91350 | |
| }, | |
| { | |
| "epoch": 26.600838379133673, | |
| "grad_norm": 0.4079902768135071, | |
| "learning_rate": 0.00028095864880605705, | |
| "loss": 3.1794, | |
| "step": 91400 | |
| }, | |
| { | |
| "epoch": 26.61539357242664, | |
| "grad_norm": 0.4160245358943939, | |
| "learning_rate": 0.00028078392545136864, | |
| "loss": 3.1706, | |
| "step": 91450 | |
| }, | |
| { | |
| "epoch": 26.629948765719607, | |
| "grad_norm": 0.41985008120536804, | |
| "learning_rate": 0.00028060920209668023, | |
| "loss": 3.1617, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 26.644503959012575, | |
| "grad_norm": 0.4122968316078186, | |
| "learning_rate": 0.00028043447874199183, | |
| "loss": 3.1802, | |
| "step": 91550 | |
| }, | |
| { | |
| "epoch": 26.65905915230554, | |
| "grad_norm": 0.4187205135822296, | |
| "learning_rate": 0.0002802597553873034, | |
| "loss": 3.1762, | |
| "step": 91600 | |
| }, | |
| { | |
| "epoch": 26.67361434559851, | |
| "grad_norm": 0.42404401302337646, | |
| "learning_rate": 0.000280085032032615, | |
| "loss": 3.1836, | |
| "step": 91650 | |
| }, | |
| { | |
| "epoch": 26.688169538891476, | |
| "grad_norm": 0.4256255328655243, | |
| "learning_rate": 0.00027991030867792656, | |
| "loss": 3.1857, | |
| "step": 91700 | |
| }, | |
| { | |
| "epoch": 26.702724732184443, | |
| "grad_norm": 0.4216116666793823, | |
| "learning_rate": 0.0002797355853232382, | |
| "loss": 3.1809, | |
| "step": 91750 | |
| }, | |
| { | |
| "epoch": 26.71727992547741, | |
| "grad_norm": 0.4425632655620575, | |
| "learning_rate": 0.00027956086196854975, | |
| "loss": 3.1756, | |
| "step": 91800 | |
| }, | |
| { | |
| "epoch": 26.731835118770377, | |
| "grad_norm": 0.4521643817424774, | |
| "learning_rate": 0.0002793861386138614, | |
| "loss": 3.1813, | |
| "step": 91850 | |
| }, | |
| { | |
| "epoch": 26.746390312063344, | |
| "grad_norm": 0.4084171950817108, | |
| "learning_rate": 0.00027921141525917293, | |
| "loss": 3.1666, | |
| "step": 91900 | |
| }, | |
| { | |
| "epoch": 26.76094550535631, | |
| "grad_norm": 0.4173433780670166, | |
| "learning_rate": 0.00027903669190448453, | |
| "loss": 3.1749, | |
| "step": 91950 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "grad_norm": 0.39873674511909485, | |
| "learning_rate": 0.0002788619685497961, | |
| "loss": 3.1964, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.775500698649278, | |
| "eval_accuracy": 0.375391263090632, | |
| "eval_loss": 3.5339925289154053, | |
| "eval_runtime": 83.9033, | |
| "eval_samples_per_second": 198.455, | |
| "eval_steps_per_second": 12.407, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 26.790055891942245, | |
| "grad_norm": 0.4373931288719177, | |
| "learning_rate": 0.0002786872451951077, | |
| "loss": 3.1718, | |
| "step": 92050 | |
| }, | |
| { | |
| "epoch": 26.804611085235212, | |
| "grad_norm": 0.4429689943790436, | |
| "learning_rate": 0.0002785125218404193, | |
| "loss": 3.1755, | |
| "step": 92100 | |
| }, | |
| { | |
| "epoch": 26.81916627852818, | |
| "grad_norm": 0.41347458958625793, | |
| "learning_rate": 0.0002783377984857309, | |
| "loss": 3.184, | |
| "step": 92150 | |
| }, | |
| { | |
| "epoch": 26.833721471821146, | |
| "grad_norm": 0.4196975529193878, | |
| "learning_rate": 0.0002781630751310425, | |
| "loss": 3.1787, | |
| "step": 92200 | |
| }, | |
| { | |
| "epoch": 26.848276665114113, | |
| "grad_norm": 0.4531739354133606, | |
| "learning_rate": 0.0002779883517763541, | |
| "loss": 3.1927, | |
| "step": 92250 | |
| }, | |
| { | |
| "epoch": 26.86283185840708, | |
| "grad_norm": 0.4349420368671417, | |
| "learning_rate": 0.0002778136284216657, | |
| "loss": 3.1853, | |
| "step": 92300 | |
| }, | |
| { | |
| "epoch": 26.877387051700047, | |
| "grad_norm": 0.41220906376838684, | |
| "learning_rate": 0.0002776389050669773, | |
| "loss": 3.1851, | |
| "step": 92350 | |
| }, | |
| { | |
| "epoch": 26.891942244993015, | |
| "grad_norm": 0.42682144045829773, | |
| "learning_rate": 0.0002774641817122888, | |
| "loss": 3.1722, | |
| "step": 92400 | |
| }, | |
| { | |
| "epoch": 26.90649743828598, | |
| "grad_norm": 0.4611343741416931, | |
| "learning_rate": 0.00027728945835760047, | |
| "loss": 3.1926, | |
| "step": 92450 | |
| }, | |
| { | |
| "epoch": 26.92105263157895, | |
| "grad_norm": 0.40907973051071167, | |
| "learning_rate": 0.000277114735002912, | |
| "loss": 3.1927, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 26.935607824871916, | |
| "grad_norm": 0.4095006287097931, | |
| "learning_rate": 0.00027694001164822366, | |
| "loss": 3.1841, | |
| "step": 92550 | |
| }, | |
| { | |
| "epoch": 26.950163018164883, | |
| "grad_norm": 0.44736406207084656, | |
| "learning_rate": 0.0002767652882935352, | |
| "loss": 3.1916, | |
| "step": 92600 | |
| }, | |
| { | |
| "epoch": 26.96471821145785, | |
| "grad_norm": 0.424541711807251, | |
| "learning_rate": 0.0002765905649388468, | |
| "loss": 3.193, | |
| "step": 92650 | |
| }, | |
| { | |
| "epoch": 26.979273404750813, | |
| "grad_norm": 0.4268820583820343, | |
| "learning_rate": 0.0002764158415841584, | |
| "loss": 3.1858, | |
| "step": 92700 | |
| }, | |
| { | |
| "epoch": 26.99382859804378, | |
| "grad_norm": 0.40377387404441833, | |
| "learning_rate": 0.00027624111822947, | |
| "loss": 3.2031, | |
| "step": 92750 | |
| }, | |
| { | |
| "epoch": 27.008150908244062, | |
| "grad_norm": 0.4540480673313141, | |
| "learning_rate": 0.0002760663948747816, | |
| "loss": 3.1316, | |
| "step": 92800 | |
| }, | |
| { | |
| "epoch": 27.02270610153703, | |
| "grad_norm": 0.41778895258903503, | |
| "learning_rate": 0.00027589167152009317, | |
| "loss": 3.1048, | |
| "step": 92850 | |
| }, | |
| { | |
| "epoch": 27.037261294829996, | |
| "grad_norm": 0.4381296634674072, | |
| "learning_rate": 0.00027571694816540477, | |
| "loss": 3.1055, | |
| "step": 92900 | |
| }, | |
| { | |
| "epoch": 27.051816488122963, | |
| "grad_norm": 0.4471452236175537, | |
| "learning_rate": 0.0002755422248107163, | |
| "loss": 3.0928, | |
| "step": 92950 | |
| }, | |
| { | |
| "epoch": 27.06637168141593, | |
| "grad_norm": 0.41887909173965454, | |
| "learning_rate": 0.00027536750145602795, | |
| "loss": 3.1036, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.06637168141593, | |
| "eval_accuracy": 0.37446597061511794, | |
| "eval_loss": 3.5499444007873535, | |
| "eval_runtime": 84.0127, | |
| "eval_samples_per_second": 198.196, | |
| "eval_steps_per_second": 12.391, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 27.080926874708897, | |
| "grad_norm": 0.43141117691993713, | |
| "learning_rate": 0.0002751927781013395, | |
| "loss": 3.1051, | |
| "step": 93050 | |
| }, | |
| { | |
| "epoch": 27.095482068001864, | |
| "grad_norm": 0.4345942735671997, | |
| "learning_rate": 0.0002750180547466511, | |
| "loss": 3.1054, | |
| "step": 93100 | |
| }, | |
| { | |
| "epoch": 27.11003726129483, | |
| "grad_norm": 0.42487776279449463, | |
| "learning_rate": 0.0002748433313919627, | |
| "loss": 3.1137, | |
| "step": 93150 | |
| }, | |
| { | |
| "epoch": 27.1245924545878, | |
| "grad_norm": 0.4264744818210602, | |
| "learning_rate": 0.0002746686080372743, | |
| "loss": 3.1135, | |
| "step": 93200 | |
| }, | |
| { | |
| "epoch": 27.139147647880765, | |
| "grad_norm": 0.4463901221752167, | |
| "learning_rate": 0.00027449388468258587, | |
| "loss": 3.1191, | |
| "step": 93250 | |
| }, | |
| { | |
| "epoch": 27.153702841173732, | |
| "grad_norm": 0.4491443336009979, | |
| "learning_rate": 0.00027431916132789747, | |
| "loss": 3.1209, | |
| "step": 93300 | |
| }, | |
| { | |
| "epoch": 27.168258034466696, | |
| "grad_norm": 0.4366472065448761, | |
| "learning_rate": 0.00027414443797320906, | |
| "loss": 3.1234, | |
| "step": 93350 | |
| }, | |
| { | |
| "epoch": 27.182813227759663, | |
| "grad_norm": 0.41730257868766785, | |
| "learning_rate": 0.00027396971461852065, | |
| "loss": 3.1316, | |
| "step": 93400 | |
| }, | |
| { | |
| "epoch": 27.19736842105263, | |
| "grad_norm": 0.4127483069896698, | |
| "learning_rate": 0.00027379499126383225, | |
| "loss": 3.1133, | |
| "step": 93450 | |
| }, | |
| { | |
| "epoch": 27.211923614345597, | |
| "grad_norm": 0.41417160630226135, | |
| "learning_rate": 0.00027362026790914384, | |
| "loss": 3.1233, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 27.226478807638564, | |
| "grad_norm": 0.3942335247993469, | |
| "learning_rate": 0.00027344554455445544, | |
| "loss": 3.131, | |
| "step": 93550 | |
| }, | |
| { | |
| "epoch": 27.24103400093153, | |
| "grad_norm": 0.4360142648220062, | |
| "learning_rate": 0.00027327082119976703, | |
| "loss": 3.1334, | |
| "step": 93600 | |
| }, | |
| { | |
| "epoch": 27.2555891942245, | |
| "grad_norm": 0.46060872077941895, | |
| "learning_rate": 0.00027309609784507857, | |
| "loss": 3.1242, | |
| "step": 93650 | |
| }, | |
| { | |
| "epoch": 27.270144387517465, | |
| "grad_norm": 0.46852752566337585, | |
| "learning_rate": 0.0002729213744903902, | |
| "loss": 3.1376, | |
| "step": 93700 | |
| }, | |
| { | |
| "epoch": 27.284699580810432, | |
| "grad_norm": 0.4117545783519745, | |
| "learning_rate": 0.00027274665113570176, | |
| "loss": 3.137, | |
| "step": 93750 | |
| }, | |
| { | |
| "epoch": 27.2992547741034, | |
| "grad_norm": 0.4287949502468109, | |
| "learning_rate": 0.00027257192778101335, | |
| "loss": 3.1369, | |
| "step": 93800 | |
| }, | |
| { | |
| "epoch": 27.313809967396367, | |
| "grad_norm": 0.4452556073665619, | |
| "learning_rate": 0.00027239720442632495, | |
| "loss": 3.1337, | |
| "step": 93850 | |
| }, | |
| { | |
| "epoch": 27.328365160689334, | |
| "grad_norm": 0.48239201307296753, | |
| "learning_rate": 0.00027222248107163654, | |
| "loss": 3.1468, | |
| "step": 93900 | |
| }, | |
| { | |
| "epoch": 27.3429203539823, | |
| "grad_norm": 0.42652374505996704, | |
| "learning_rate": 0.00027204775771694814, | |
| "loss": 3.1443, | |
| "step": 93950 | |
| }, | |
| { | |
| "epoch": 27.357475547275268, | |
| "grad_norm": 0.45868149399757385, | |
| "learning_rate": 0.00027187303436225973, | |
| "loss": 3.1444, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.357475547275268, | |
| "eval_accuracy": 0.3749929630525884, | |
| "eval_loss": 3.542569160461426, | |
| "eval_runtime": 83.92, | |
| "eval_samples_per_second": 198.415, | |
| "eval_steps_per_second": 12.405, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 27.372030740568235, | |
| "grad_norm": 0.421965092420578, | |
| "learning_rate": 0.0002716983110075713, | |
| "loss": 3.1372, | |
| "step": 94050 | |
| }, | |
| { | |
| "epoch": 27.386585933861202, | |
| "grad_norm": 0.43176019191741943, | |
| "learning_rate": 0.0002715235876528829, | |
| "loss": 3.1322, | |
| "step": 94100 | |
| }, | |
| { | |
| "epoch": 27.40114112715417, | |
| "grad_norm": 0.43462324142456055, | |
| "learning_rate": 0.0002713488642981945, | |
| "loss": 3.14, | |
| "step": 94150 | |
| }, | |
| { | |
| "epoch": 27.415696320447136, | |
| "grad_norm": 0.4510142505168915, | |
| "learning_rate": 0.00027117414094350606, | |
| "loss": 3.1498, | |
| "step": 94200 | |
| }, | |
| { | |
| "epoch": 27.430251513740103, | |
| "grad_norm": 0.4283021092414856, | |
| "learning_rate": 0.0002709994175888177, | |
| "loss": 3.1429, | |
| "step": 94250 | |
| }, | |
| { | |
| "epoch": 27.44480670703307, | |
| "grad_norm": 0.41187649965286255, | |
| "learning_rate": 0.00027082469423412924, | |
| "loss": 3.1525, | |
| "step": 94300 | |
| }, | |
| { | |
| "epoch": 27.459361900326037, | |
| "grad_norm": 0.5103968977928162, | |
| "learning_rate": 0.00027064997087944084, | |
| "loss": 3.1492, | |
| "step": 94350 | |
| }, | |
| { | |
| "epoch": 27.473917093619004, | |
| "grad_norm": 0.43091094493865967, | |
| "learning_rate": 0.00027047524752475243, | |
| "loss": 3.1596, | |
| "step": 94400 | |
| }, | |
| { | |
| "epoch": 27.48847228691197, | |
| "grad_norm": 0.4661984443664551, | |
| "learning_rate": 0.000270300524170064, | |
| "loss": 3.1567, | |
| "step": 94450 | |
| }, | |
| { | |
| "epoch": 27.50302748020494, | |
| "grad_norm": 0.43993762135505676, | |
| "learning_rate": 0.0002701258008153757, | |
| "loss": 3.1562, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 27.517582673497905, | |
| "grad_norm": 0.4352594017982483, | |
| "learning_rate": 0.0002699510774606872, | |
| "loss": 3.158, | |
| "step": 94550 | |
| }, | |
| { | |
| "epoch": 27.532137866790872, | |
| "grad_norm": 0.41682329773902893, | |
| "learning_rate": 0.0002697763541059988, | |
| "loss": 3.1598, | |
| "step": 94600 | |
| }, | |
| { | |
| "epoch": 27.54669306008384, | |
| "grad_norm": 0.420386403799057, | |
| "learning_rate": 0.0002696016307513104, | |
| "loss": 3.1541, | |
| "step": 94650 | |
| }, | |
| { | |
| "epoch": 27.561248253376803, | |
| "grad_norm": 0.46121206879615784, | |
| "learning_rate": 0.000269426907396622, | |
| "loss": 3.1571, | |
| "step": 94700 | |
| }, | |
| { | |
| "epoch": 27.57580344666977, | |
| "grad_norm": 0.4396849572658539, | |
| "learning_rate": 0.0002692521840419336, | |
| "loss": 3.1633, | |
| "step": 94750 | |
| }, | |
| { | |
| "epoch": 27.590358639962737, | |
| "grad_norm": 0.42374372482299805, | |
| "learning_rate": 0.0002690774606872452, | |
| "loss": 3.1471, | |
| "step": 94800 | |
| }, | |
| { | |
| "epoch": 27.604913833255704, | |
| "grad_norm": 0.47300276160240173, | |
| "learning_rate": 0.0002689027373325568, | |
| "loss": 3.1589, | |
| "step": 94850 | |
| }, | |
| { | |
| "epoch": 27.61946902654867, | |
| "grad_norm": 0.4443678855895996, | |
| "learning_rate": 0.0002687280139778683, | |
| "loss": 3.1715, | |
| "step": 94900 | |
| }, | |
| { | |
| "epoch": 27.63402421984164, | |
| "grad_norm": 0.44060665369033813, | |
| "learning_rate": 0.00026855329062317997, | |
| "loss": 3.1734, | |
| "step": 94950 | |
| }, | |
| { | |
| "epoch": 27.648579413134605, | |
| "grad_norm": 0.4386674165725708, | |
| "learning_rate": 0.0002683785672684915, | |
| "loss": 3.1646, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.648579413134605, | |
| "eval_accuracy": 0.37534824809685097, | |
| "eval_loss": 3.5367236137390137, | |
| "eval_runtime": 83.9898, | |
| "eval_samples_per_second": 198.25, | |
| "eval_steps_per_second": 12.394, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 27.663134606427572, | |
| "grad_norm": 0.44010964035987854, | |
| "learning_rate": 0.0002682038439138031, | |
| "loss": 3.1852, | |
| "step": 95050 | |
| }, | |
| { | |
| "epoch": 27.67768979972054, | |
| "grad_norm": 0.4325070381164551, | |
| "learning_rate": 0.0002680291205591147, | |
| "loss": 3.1812, | |
| "step": 95100 | |
| }, | |
| { | |
| "epoch": 27.692244993013507, | |
| "grad_norm": 0.42794081568717957, | |
| "learning_rate": 0.0002678543972044263, | |
| "loss": 3.1695, | |
| "step": 95150 | |
| }, | |
| { | |
| "epoch": 27.706800186306474, | |
| "grad_norm": 0.4375181496143341, | |
| "learning_rate": 0.0002676796738497379, | |
| "loss": 3.1701, | |
| "step": 95200 | |
| }, | |
| { | |
| "epoch": 27.72135537959944, | |
| "grad_norm": 0.4309106469154358, | |
| "learning_rate": 0.0002675049504950495, | |
| "loss": 3.1632, | |
| "step": 95250 | |
| }, | |
| { | |
| "epoch": 27.735910572892408, | |
| "grad_norm": 0.4378839135169983, | |
| "learning_rate": 0.0002673302271403611, | |
| "loss": 3.1652, | |
| "step": 95300 | |
| }, | |
| { | |
| "epoch": 27.750465766185375, | |
| "grad_norm": 0.4284781217575073, | |
| "learning_rate": 0.00026715550378567267, | |
| "loss": 3.1689, | |
| "step": 95350 | |
| }, | |
| { | |
| "epoch": 27.765020959478342, | |
| "grad_norm": 0.4057956039905548, | |
| "learning_rate": 0.00026698078043098426, | |
| "loss": 3.1633, | |
| "step": 95400 | |
| }, | |
| { | |
| "epoch": 27.77957615277131, | |
| "grad_norm": 0.4229182302951813, | |
| "learning_rate": 0.00026680605707629586, | |
| "loss": 3.1712, | |
| "step": 95450 | |
| }, | |
| { | |
| "epoch": 27.794131346064276, | |
| "grad_norm": 0.4494783878326416, | |
| "learning_rate": 0.00026663133372160745, | |
| "loss": 3.1732, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 27.808686539357243, | |
| "grad_norm": 0.436646431684494, | |
| "learning_rate": 0.00026645661036691905, | |
| "loss": 3.1747, | |
| "step": 95550 | |
| }, | |
| { | |
| "epoch": 27.82324173265021, | |
| "grad_norm": 0.44214946031570435, | |
| "learning_rate": 0.0002662818870122306, | |
| "loss": 3.1786, | |
| "step": 95600 | |
| }, | |
| { | |
| "epoch": 27.837796925943177, | |
| "grad_norm": 0.4244590699672699, | |
| "learning_rate": 0.00026610716365754224, | |
| "loss": 3.1768, | |
| "step": 95650 | |
| }, | |
| { | |
| "epoch": 27.852352119236144, | |
| "grad_norm": 0.430711954832077, | |
| "learning_rate": 0.0002659324403028538, | |
| "loss": 3.1799, | |
| "step": 95700 | |
| }, | |
| { | |
| "epoch": 27.86690731252911, | |
| "grad_norm": 0.4592069685459137, | |
| "learning_rate": 0.00026575771694816537, | |
| "loss": 3.1844, | |
| "step": 95750 | |
| }, | |
| { | |
| "epoch": 27.88146250582208, | |
| "grad_norm": 0.4398399591445923, | |
| "learning_rate": 0.00026558299359347696, | |
| "loss": 3.1858, | |
| "step": 95800 | |
| }, | |
| { | |
| "epoch": 27.896017699115045, | |
| "grad_norm": 0.458998441696167, | |
| "learning_rate": 0.00026540827023878856, | |
| "loss": 3.1719, | |
| "step": 95850 | |
| }, | |
| { | |
| "epoch": 27.910572892408013, | |
| "grad_norm": 0.4320071041584015, | |
| "learning_rate": 0.00026523354688410015, | |
| "loss": 3.1713, | |
| "step": 95900 | |
| }, | |
| { | |
| "epoch": 27.92512808570098, | |
| "grad_norm": 0.44292187690734863, | |
| "learning_rate": 0.00026505882352941175, | |
| "loss": 3.1729, | |
| "step": 95950 | |
| }, | |
| { | |
| "epoch": 27.939683278993947, | |
| "grad_norm": 0.45965898036956787, | |
| "learning_rate": 0.00026488410017472334, | |
| "loss": 3.1847, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.939683278993947, | |
| "eval_accuracy": 0.37562725791990065, | |
| "eval_loss": 3.531078815460205, | |
| "eval_runtime": 84.0859, | |
| "eval_samples_per_second": 198.024, | |
| "eval_steps_per_second": 12.38, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 27.95423847228691, | |
| "grad_norm": 0.4116367697715759, | |
| "learning_rate": 0.00026470937682003494, | |
| "loss": 3.1874, | |
| "step": 96050 | |
| }, | |
| { | |
| "epoch": 27.968793665579877, | |
| "grad_norm": 0.43813031911849976, | |
| "learning_rate": 0.00026453465346534653, | |
| "loss": 3.1809, | |
| "step": 96100 | |
| }, | |
| { | |
| "epoch": 27.983348858872844, | |
| "grad_norm": 0.42912182211875916, | |
| "learning_rate": 0.00026435993011065807, | |
| "loss": 3.1865, | |
| "step": 96150 | |
| }, | |
| { | |
| "epoch": 27.99790405216581, | |
| "grad_norm": 0.4206131100654602, | |
| "learning_rate": 0.0002641852067559697, | |
| "loss": 3.1743, | |
| "step": 96200 | |
| }, | |
| { | |
| "epoch": 28.012226362366093, | |
| "grad_norm": 0.42363131046295166, | |
| "learning_rate": 0.00026401048340128126, | |
| "loss": 3.1125, | |
| "step": 96250 | |
| }, | |
| { | |
| "epoch": 28.02678155565906, | |
| "grad_norm": 0.45002052187919617, | |
| "learning_rate": 0.00026383576004659285, | |
| "loss": 3.0914, | |
| "step": 96300 | |
| }, | |
| { | |
| "epoch": 28.041336748952027, | |
| "grad_norm": 0.434286504983902, | |
| "learning_rate": 0.00026366103669190445, | |
| "loss": 3.0921, | |
| "step": 96350 | |
| }, | |
| { | |
| "epoch": 28.055891942244994, | |
| "grad_norm": 0.4626716077327728, | |
| "learning_rate": 0.00026348631333721604, | |
| "loss": 3.0875, | |
| "step": 96400 | |
| }, | |
| { | |
| "epoch": 28.07044713553796, | |
| "grad_norm": 0.4696536064147949, | |
| "learning_rate": 0.00026331158998252764, | |
| "loss": 3.0986, | |
| "step": 96450 | |
| }, | |
| { | |
| "epoch": 28.085002328830928, | |
| "grad_norm": 0.431812584400177, | |
| "learning_rate": 0.00026313686662783923, | |
| "loss": 3.097, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 28.099557522123895, | |
| "grad_norm": 0.40847593545913696, | |
| "learning_rate": 0.0002629621432731508, | |
| "loss": 3.1014, | |
| "step": 96550 | |
| }, | |
| { | |
| "epoch": 28.114112715416862, | |
| "grad_norm": 0.4362127482891083, | |
| "learning_rate": 0.0002627874199184624, | |
| "loss": 3.1084, | |
| "step": 96600 | |
| }, | |
| { | |
| "epoch": 28.12866790870983, | |
| "grad_norm": 0.46820372343063354, | |
| "learning_rate": 0.000262612696563774, | |
| "loss": 3.0987, | |
| "step": 96650 | |
| }, | |
| { | |
| "epoch": 28.143223102002796, | |
| "grad_norm": 0.47318655252456665, | |
| "learning_rate": 0.0002624379732090856, | |
| "loss": 3.1023, | |
| "step": 96700 | |
| }, | |
| { | |
| "epoch": 28.15777829529576, | |
| "grad_norm": 0.43643152713775635, | |
| "learning_rate": 0.0002622632498543972, | |
| "loss": 3.1083, | |
| "step": 96750 | |
| }, | |
| { | |
| "epoch": 28.172333488588727, | |
| "grad_norm": 0.4407923221588135, | |
| "learning_rate": 0.0002620885264997088, | |
| "loss": 3.1137, | |
| "step": 96800 | |
| }, | |
| { | |
| "epoch": 28.186888681881694, | |
| "grad_norm": 0.4425899386405945, | |
| "learning_rate": 0.00026191380314502034, | |
| "loss": 3.1157, | |
| "step": 96850 | |
| }, | |
| { | |
| "epoch": 28.20144387517466, | |
| "grad_norm": 0.42267611622810364, | |
| "learning_rate": 0.000261739079790332, | |
| "loss": 3.1126, | |
| "step": 96900 | |
| }, | |
| { | |
| "epoch": 28.215999068467628, | |
| "grad_norm": 0.44764307141304016, | |
| "learning_rate": 0.0002615643564356435, | |
| "loss": 3.1248, | |
| "step": 96950 | |
| }, | |
| { | |
| "epoch": 28.230554261760595, | |
| "grad_norm": 0.42632830142974854, | |
| "learning_rate": 0.0002613896330809551, | |
| "loss": 3.1124, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.230554261760595, | |
| "eval_accuracy": 0.3749954311260021, | |
| "eval_loss": 3.5469326972961426, | |
| "eval_runtime": 84.0215, | |
| "eval_samples_per_second": 198.176, | |
| "eval_steps_per_second": 12.39, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 28.245109455053562, | |
| "grad_norm": 0.4517412781715393, | |
| "learning_rate": 0.0002612149097262667, | |
| "loss": 3.1202, | |
| "step": 97050 | |
| }, | |
| { | |
| "epoch": 28.25966464834653, | |
| "grad_norm": 0.44456765055656433, | |
| "learning_rate": 0.0002610401863715783, | |
| "loss": 3.1323, | |
| "step": 97100 | |
| }, | |
| { | |
| "epoch": 28.274219841639496, | |
| "grad_norm": 0.4247918725013733, | |
| "learning_rate": 0.0002608654630168899, | |
| "loss": 3.1362, | |
| "step": 97150 | |
| }, | |
| { | |
| "epoch": 28.288775034932463, | |
| "grad_norm": 0.501080334186554, | |
| "learning_rate": 0.0002606907396622015, | |
| "loss": 3.1246, | |
| "step": 97200 | |
| }, | |
| { | |
| "epoch": 28.30333022822543, | |
| "grad_norm": 0.4519060254096985, | |
| "learning_rate": 0.0002605160163075131, | |
| "loss": 3.1388, | |
| "step": 97250 | |
| }, | |
| { | |
| "epoch": 28.317885421518397, | |
| "grad_norm": 0.43544185161590576, | |
| "learning_rate": 0.0002603412929528247, | |
| "loss": 3.1296, | |
| "step": 97300 | |
| }, | |
| { | |
| "epoch": 28.332440614811365, | |
| "grad_norm": 0.4498087167739868, | |
| "learning_rate": 0.0002601665695981363, | |
| "loss": 3.1278, | |
| "step": 97350 | |
| }, | |
| { | |
| "epoch": 28.34699580810433, | |
| "grad_norm": 0.44666847586631775, | |
| "learning_rate": 0.0002599918462434478, | |
| "loss": 3.1463, | |
| "step": 97400 | |
| }, | |
| { | |
| "epoch": 28.3615510013973, | |
| "grad_norm": 0.44268566370010376, | |
| "learning_rate": 0.00025981712288875947, | |
| "loss": 3.1418, | |
| "step": 97450 | |
| }, | |
| { | |
| "epoch": 28.376106194690266, | |
| "grad_norm": 0.4413694441318512, | |
| "learning_rate": 0.000259642399534071, | |
| "loss": 3.1294, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 28.390661387983233, | |
| "grad_norm": 0.4524979889392853, | |
| "learning_rate": 0.0002594676761793826, | |
| "loss": 3.1361, | |
| "step": 97550 | |
| }, | |
| { | |
| "epoch": 28.4052165812762, | |
| "grad_norm": 0.4177760183811188, | |
| "learning_rate": 0.0002592929528246942, | |
| "loss": 3.1383, | |
| "step": 97600 | |
| }, | |
| { | |
| "epoch": 28.419771774569167, | |
| "grad_norm": 0.4174005687236786, | |
| "learning_rate": 0.0002591182294700058, | |
| "loss": 3.1331, | |
| "step": 97650 | |
| }, | |
| { | |
| "epoch": 28.434326967862134, | |
| "grad_norm": 0.42065516114234924, | |
| "learning_rate": 0.0002589435061153174, | |
| "loss": 3.142, | |
| "step": 97700 | |
| }, | |
| { | |
| "epoch": 28.4488821611551, | |
| "grad_norm": 0.46468207240104675, | |
| "learning_rate": 0.000258768782760629, | |
| "loss": 3.1419, | |
| "step": 97750 | |
| }, | |
| { | |
| "epoch": 28.463437354448068, | |
| "grad_norm": 0.4186297357082367, | |
| "learning_rate": 0.0002585940594059406, | |
| "loss": 3.1375, | |
| "step": 97800 | |
| }, | |
| { | |
| "epoch": 28.477992547741035, | |
| "grad_norm": 0.49948257207870483, | |
| "learning_rate": 0.00025841933605125217, | |
| "loss": 3.1527, | |
| "step": 97850 | |
| }, | |
| { | |
| "epoch": 28.492547741034002, | |
| "grad_norm": 0.4354049265384674, | |
| "learning_rate": 0.00025824461269656376, | |
| "loss": 3.148, | |
| "step": 97900 | |
| }, | |
| { | |
| "epoch": 28.50710293432697, | |
| "grad_norm": 0.42570626735687256, | |
| "learning_rate": 0.00025806988934187536, | |
| "loss": 3.1614, | |
| "step": 97950 | |
| }, | |
| { | |
| "epoch": 28.521658127619936, | |
| "grad_norm": 0.4368693232536316, | |
| "learning_rate": 0.00025789516598718695, | |
| "loss": 3.1577, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.521658127619936, | |
| "eval_accuracy": 0.3751945223813712, | |
| "eval_loss": 3.539001941680908, | |
| "eval_runtime": 84.0442, | |
| "eval_samples_per_second": 198.122, | |
| "eval_steps_per_second": 12.386, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 28.536213320912903, | |
| "grad_norm": 0.45667126774787903, | |
| "learning_rate": 0.00025772044263249854, | |
| "loss": 3.1501, | |
| "step": 98050 | |
| }, | |
| { | |
| "epoch": 28.55076851420587, | |
| "grad_norm": 0.4505082964897156, | |
| "learning_rate": 0.0002575457192778101, | |
| "loss": 3.1427, | |
| "step": 98100 | |
| }, | |
| { | |
| "epoch": 28.565323707498834, | |
| "grad_norm": 0.4331296980381012, | |
| "learning_rate": 0.00025737099592312173, | |
| "loss": 3.1517, | |
| "step": 98150 | |
| }, | |
| { | |
| "epoch": 28.5798789007918, | |
| "grad_norm": 0.44055548310279846, | |
| "learning_rate": 0.0002571962725684333, | |
| "loss": 3.146, | |
| "step": 98200 | |
| }, | |
| { | |
| "epoch": 28.594434094084768, | |
| "grad_norm": 0.46562591195106506, | |
| "learning_rate": 0.00025702154921374487, | |
| "loss": 3.1569, | |
| "step": 98250 | |
| }, | |
| { | |
| "epoch": 28.608989287377735, | |
| "grad_norm": 0.4513644278049469, | |
| "learning_rate": 0.00025684682585905646, | |
| "loss": 3.1518, | |
| "step": 98300 | |
| }, | |
| { | |
| "epoch": 28.623544480670702, | |
| "grad_norm": 0.4585248529911041, | |
| "learning_rate": 0.00025667210250436806, | |
| "loss": 3.1596, | |
| "step": 98350 | |
| }, | |
| { | |
| "epoch": 28.63809967396367, | |
| "grad_norm": 0.46808040142059326, | |
| "learning_rate": 0.00025649737914967965, | |
| "loss": 3.1527, | |
| "step": 98400 | |
| }, | |
| { | |
| "epoch": 28.652654867256636, | |
| "grad_norm": 0.4026799201965332, | |
| "learning_rate": 0.00025632265579499124, | |
| "loss": 3.1584, | |
| "step": 98450 | |
| }, | |
| { | |
| "epoch": 28.667210060549603, | |
| "grad_norm": 0.44900086522102356, | |
| "learning_rate": 0.00025614793244030284, | |
| "loss": 3.1601, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 28.68176525384257, | |
| "grad_norm": 0.44888389110565186, | |
| "learning_rate": 0.0002559732090856144, | |
| "loss": 3.1644, | |
| "step": 98550 | |
| }, | |
| { | |
| "epoch": 28.696320447135538, | |
| "grad_norm": 0.44954290986061096, | |
| "learning_rate": 0.00025579848573092603, | |
| "loss": 3.1647, | |
| "step": 98600 | |
| }, | |
| { | |
| "epoch": 28.710875640428505, | |
| "grad_norm": 0.4544377028942108, | |
| "learning_rate": 0.0002556237623762376, | |
| "loss": 3.1671, | |
| "step": 98650 | |
| }, | |
| { | |
| "epoch": 28.72543083372147, | |
| "grad_norm": 0.41961169242858887, | |
| "learning_rate": 0.0002554490390215492, | |
| "loss": 3.1552, | |
| "step": 98700 | |
| }, | |
| { | |
| "epoch": 28.73998602701444, | |
| "grad_norm": 0.4601042866706848, | |
| "learning_rate": 0.0002552743156668608, | |
| "loss": 3.1561, | |
| "step": 98750 | |
| }, | |
| { | |
| "epoch": 28.754541220307406, | |
| "grad_norm": 0.45733726024627686, | |
| "learning_rate": 0.00025509959231217235, | |
| "loss": 3.1537, | |
| "step": 98800 | |
| }, | |
| { | |
| "epoch": 28.769096413600373, | |
| "grad_norm": 0.43923458456993103, | |
| "learning_rate": 0.000254924868957484, | |
| "loss": 3.1596, | |
| "step": 98850 | |
| }, | |
| { | |
| "epoch": 28.78365160689334, | |
| "grad_norm": 0.4651845097541809, | |
| "learning_rate": 0.00025475014560279554, | |
| "loss": 3.1652, | |
| "step": 98900 | |
| }, | |
| { | |
| "epoch": 28.798206800186307, | |
| "grad_norm": 0.43512633442878723, | |
| "learning_rate": 0.00025457542224810713, | |
| "loss": 3.1578, | |
| "step": 98950 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "grad_norm": 0.4440561532974243, | |
| "learning_rate": 0.00025440069889341873, | |
| "loss": 3.1592, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "eval_accuracy": 0.375726450965669, | |
| "eval_loss": 3.5344488620758057, | |
| "eval_runtime": 84.1833, | |
| "eval_samples_per_second": 197.795, | |
| "eval_steps_per_second": 12.366, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 28.812761993479274, | |
| "step": 99000, | |
| "total_flos": 2.068898812526592e+18, | |
| "train_loss": 3.3872425547898417, | |
| "train_runtime": 73941.0651, | |
| "train_samples_per_second": 185.828, | |
| "train_steps_per_second": 2.323 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 171800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 20, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 11 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.068898812526592e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |