| { |
| "best_global_step": 1000, |
| "best_metric": 0.988082230091095, |
| "best_model_checkpoint": "./sft-out/checkpoint-1000", |
| "epoch": 5.3005690922150395, |
| "eval_steps": 500, |
| "global_step": 3000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.08840267418089397, |
| "grad_norm": 2.867664337158203, |
| "learning_rate": 9.800000000000001e-06, |
| "loss": 1.6486582946777344, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.17680534836178793, |
| "grad_norm": 2.999215841293335, |
| "learning_rate": 1.98e-05, |
| "loss": 1.3320436096191406, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.26520802254268194, |
| "grad_norm": 2.5709331035614014, |
| "learning_rate": 1.998591477024944e-05, |
| "loss": 1.2357231903076171, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.35361069672357587, |
| "grad_norm": 2.5824177265167236, |
| "learning_rate": 1.994254499970557e-05, |
| "loss": 1.2124821472167968, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.44201337090446985, |
| "grad_norm": 2.1276214122772217, |
| "learning_rate": 1.987001203159941e-05, |
| "loss": 1.1784111785888671, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.5304160450853639, |
| "grad_norm": 2.2801501750946045, |
| "learning_rate": 1.9768528617623674e-05, |
| "loss": 1.14390380859375, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.6188187192662578, |
| "grad_norm": 2.245532989501953, |
| "learning_rate": 1.963839242611693e-05, |
| "loss": 1.1346678161621093, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.7072213934471517, |
| "grad_norm": 2.0729880332946777, |
| "learning_rate": 1.947998516895103e-05, |
| "loss": 1.1167774963378907, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7956240676280457, |
| "grad_norm": 2.183257818222046, |
| "learning_rate": 1.9293771481904463e-05, |
| "loss": 1.0449161529541016, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.8840267418089397, |
| "grad_norm": 2.2573604583740234, |
| "learning_rate": 1.9080297561805535e-05, |
| "loss": 1.0654985809326172, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8840267418089397, |
| "eval_loss": 1.041561245918274, |
| "eval_runtime": 27.497, |
| "eval_samples_per_second": 18.184, |
| "eval_steps_per_second": 2.291, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.9724294159898337, |
| "grad_norm": 2.2783753871917725, |
| "learning_rate": 1.884018956444297e-05, |
| "loss": 1.068794708251953, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.060113818443008, |
| "grad_norm": 2.0879323482513428, |
| "learning_rate": 1.85741517679431e-05, |
| "loss": 0.8203845977783203, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.148516492623902, |
| "grad_norm": 2.290803909301758, |
| "learning_rate": 1.828296450700078e-05, |
| "loss": 0.7069022369384765, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.2369191668047959, |
| "grad_norm": 2.6409525871276855, |
| "learning_rate": 1.7967481884023257e-05, |
| "loss": 0.707320556640625, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.3253218409856897, |
| "grad_norm": 2.363614559173584, |
| "learning_rate": 1.7628629263900676e-05, |
| "loss": 0.7089004516601562, |
| "step": 750 |
| }, |
| { |
| "epoch": 1.4137245151665838, |
| "grad_norm": 2.3598053455352783, |
| "learning_rate": 1.7267400559751396e-05, |
| "loss": 0.709156265258789, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.502127189347478, |
| "grad_norm": 2.0364632606506348, |
| "learning_rate": 1.6884855317603584e-05, |
| "loss": 0.6911470031738282, |
| "step": 850 |
| }, |
| { |
| "epoch": 1.5905298635283718, |
| "grad_norm": 2.0562117099761963, |
| "learning_rate": 1.648211560856415e-05, |
| "loss": 0.6989401245117187, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.6789325377092656, |
| "grad_norm": 2.351238965988159, |
| "learning_rate": 1.6060362737590845e-05, |
| "loss": 0.7044075775146484, |
| "step": 950 |
| }, |
| { |
| "epoch": 1.7673352118901597, |
| "grad_norm": 2.1121199131011963, |
| "learning_rate": 1.5620833778521306e-05, |
| "loss": 0.7032404327392578, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7673352118901597, |
| "eval_loss": 0.988082230091095, |
| "eval_runtime": 27.3518, |
| "eval_samples_per_second": 18.28, |
| "eval_steps_per_second": 2.303, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.8557378860710536, |
| "grad_norm": 2.1198668479919434, |
| "learning_rate": 1.5164817945522343e-05, |
| "loss": 0.7200093078613281, |
| "step": 1050 |
| }, |
| { |
| "epoch": 1.9441405602519475, |
| "grad_norm": 2.416775703430176, |
| "learning_rate": 1.4693652811602635e-05, |
| "loss": 0.7036862182617187, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.0318249627051217, |
| "grad_norm": 2.5851995944976807, |
| "learning_rate": 1.4208720385280673e-05, |
| "loss": 0.5860164260864258, |
| "step": 1150 |
| }, |
| { |
| "epoch": 2.120227636886016, |
| "grad_norm": 1.8049288988113403, |
| "learning_rate": 1.3711443056915611e-05, |
| "loss": 0.354575080871582, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.20863031106691, |
| "grad_norm": 2.2025763988494873, |
| "learning_rate": 1.3203279426591329e-05, |
| "loss": 0.3392633819580078, |
| "step": 1250 |
| }, |
| { |
| "epoch": 2.297032985247804, |
| "grad_norm": 2.0318801403045654, |
| "learning_rate": 1.2685720025791004e-05, |
| "loss": 0.33905509948730467, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.3854356594286976, |
| "grad_norm": 2.0977327823638916, |
| "learning_rate": 1.2160282945411513e-05, |
| "loss": 0.34515163421630857, |
| "step": 1350 |
| }, |
| { |
| "epoch": 2.4738383336095917, |
| "grad_norm": 2.108346462249756, |
| "learning_rate": 1.1628509382941232e-05, |
| "loss": 0.35274150848388675, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.562241007790486, |
| "grad_norm": 2.778698682785034, |
| "learning_rate": 1.1091959121862279e-05, |
| "loss": 0.3534046173095703, |
| "step": 1450 |
| }, |
| { |
| "epoch": 2.6506436819713795, |
| "grad_norm": 2.0644469261169434, |
| "learning_rate": 1.0552205956536803e-05, |
| "loss": 0.34923103332519534, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.6506436819713795, |
| "eval_loss": 1.0509159564971924, |
| "eval_runtime": 27.3535, |
| "eval_samples_per_second": 18.279, |
| "eval_steps_per_second": 2.303, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.7390463561522735, |
| "grad_norm": 1.9865392446517944, |
| "learning_rate": 1.001083307599696e-05, |
| "loss": 0.34075843811035156, |
| "step": 1550 |
| }, |
| { |
| "epoch": 2.8274490303331676, |
| "grad_norm": 2.692328691482544, |
| "learning_rate": 9.469428420178699e-06, |
| "loss": 0.3529329299926758, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.9158517045140613, |
| "grad_norm": 2.0326497554779053, |
| "learning_rate": 8.92958002222035e-06, |
| "loss": 0.35016448974609377, |
| "step": 1650 |
| }, |
| { |
| "epoch": 3.003536106967236, |
| "grad_norm": 1.2595773935317993, |
| "learning_rate": 8.392871350487781e-06, |
| "loss": 0.3463246154785156, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0919387811481296, |
| "grad_norm": 1.5270416736602783, |
| "learning_rate": 7.860876663988886e-06, |
| "loss": 0.14206992149353026, |
| "step": 1750 |
| }, |
| { |
| "epoch": 3.1803414553290237, |
| "grad_norm": 1.808855652809143, |
| "learning_rate": 7.335156394800651e-06, |
| "loss": 0.14106608390808106, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.2687441295099178, |
| "grad_norm": 1.834566593170166, |
| "learning_rate": 6.817252571053034e-06, |
| "loss": 0.138691987991333, |
| "step": 1850 |
| }, |
| { |
| "epoch": 3.357146803690812, |
| "grad_norm": 1.779943823814392, |
| "learning_rate": 6.308684293894747e-06, |
| "loss": 0.13664389610290528, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.4455494778717055, |
| "grad_norm": 1.5618979930877686, |
| "learning_rate": 5.810943281707862e-06, |
| "loss": 0.13606584548950196, |
| "step": 1950 |
| }, |
| { |
| "epoch": 3.5339521520525996, |
| "grad_norm": 1.8645941019058228, |
| "learning_rate": 5.325489494640788e-06, |
| "loss": 0.13294898033142089, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.5339521520525996, |
| "eval_loss": 1.1485551595687866, |
| "eval_runtime": 27.413, |
| "eval_samples_per_second": 18.24, |
| "eval_steps_per_second": 2.298, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.6223548262334937, |
| "grad_norm": 1.6168675422668457, |
| "learning_rate": 4.853746852293582e-06, |
| "loss": 0.13220906257629395, |
| "step": 2050 |
| }, |
| { |
| "epoch": 3.7107575004143873, |
| "grad_norm": 1.501587152481079, |
| "learning_rate": 4.397099057116449e-06, |
| "loss": 0.13320664405822755, |
| "step": 2100 |
| }, |
| { |
| "epoch": 3.7991601745952814, |
| "grad_norm": 1.7080086469650269, |
| "learning_rate": 3.956885535772039e-06, |
| "loss": 0.13129110336303712, |
| "step": 2150 |
| }, |
| { |
| "epoch": 3.8875628487761755, |
| "grad_norm": 1.4526396989822388, |
| "learning_rate": 3.534397510366242e-06, |
| "loss": 0.1242832088470459, |
| "step": 2200 |
| }, |
| { |
| "epoch": 3.975965522957069, |
| "grad_norm": 1.5429104566574097, |
| "learning_rate": 3.1308742110713165e-06, |
| "loss": 0.1259861946105957, |
| "step": 2250 |
| }, |
| { |
| "epoch": 4.063649925410243, |
| "grad_norm": 0.9340503811836243, |
| "learning_rate": 2.7474992412502555e-06, |
| "loss": 0.07559285640716552, |
| "step": 2300 |
| }, |
| { |
| "epoch": 4.152052599591138, |
| "grad_norm": 1.0590145587921143, |
| "learning_rate": 2.385397105744187e-06, |
| "loss": 0.05158629417419434, |
| "step": 2350 |
| }, |
| { |
| "epoch": 4.240455273772032, |
| "grad_norm": 1.2688530683517456, |
| "learning_rate": 2.04562991250593e-06, |
| "loss": 0.0525316333770752, |
| "step": 2400 |
| }, |
| { |
| "epoch": 4.328857947952925, |
| "grad_norm": 1.6087177991867065, |
| "learning_rate": 1.7291942572543806e-06, |
| "loss": 0.049799141883850095, |
| "step": 2450 |
| }, |
| { |
| "epoch": 4.41726062213382, |
| "grad_norm": 0.8664147853851318, |
| "learning_rate": 1.437018300287546e-06, |
| "loss": 0.05318857669830322, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.41726062213382, |
| "eval_loss": 1.2274935245513916, |
| "eval_runtime": 27.4265, |
| "eval_samples_per_second": 18.231, |
| "eval_steps_per_second": 2.297, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.505663296314713, |
| "grad_norm": 1.0118080377578735, |
| "learning_rate": 1.169959044028467e-06, |
| "loss": 0.051605224609375, |
| "step": 2550 |
| }, |
| { |
| "epoch": 4.594065970495608, |
| "grad_norm": 1.1289219856262207, |
| "learning_rate": 9.287998192894598e-07, |
| "loss": 0.04904799938201904, |
| "step": 2600 |
| }, |
| { |
| "epoch": 4.682468644676502, |
| "grad_norm": 0.7907158732414246, |
| "learning_rate": 7.142479876278796e-07, |
| "loss": 0.048020644187927244, |
| "step": 2650 |
| }, |
| { |
| "epoch": 4.770871318857395, |
| "grad_norm": 0.8089121580123901, |
| "learning_rate": 5.269328665327855e-07, |
| "loss": 0.05245284557342529, |
| "step": 2700 |
| }, |
| { |
| "epoch": 4.85927399303829, |
| "grad_norm": 0.880215585231781, |
| "learning_rate": 3.6740388352832666e-07, |
| "loss": 0.04892232418060303, |
| "step": 2750 |
| }, |
| { |
| "epoch": 4.947676667219183, |
| "grad_norm": 0.9871537685394287, |
| "learning_rate": 2.3612896460813772e-07, |
| "loss": 0.04938845157623291, |
| "step": 2800 |
| }, |
| { |
| "epoch": 5.035361069672358, |
| "grad_norm": 0.5592436790466309, |
| "learning_rate": 1.334931617277857e-07, |
| "loss": 0.0440280818939209, |
| "step": 2850 |
| }, |
| { |
| "epoch": 5.123763743853251, |
| "grad_norm": 0.5856468677520752, |
| "learning_rate": 5.979752338103661e-08, |
| "loss": 0.03595466136932373, |
| "step": 2900 |
| }, |
| { |
| "epoch": 5.212166418034146, |
| "grad_norm": 0.4753524363040924, |
| "learning_rate": 1.52582115727562e-08, |
| "loss": 0.03345550060272217, |
| "step": 2950 |
| }, |
| { |
| "epoch": 5.3005690922150395, |
| "grad_norm": 0.625528872013092, |
| "learning_rate": 5.867778499757393e-12, |
| "loss": 0.03459673643112183, |
| "step": 3000 |
| }, |
| { |
| "epoch": 5.3005690922150395, |
| "eval_loss": 1.2397174835205078, |
| "eval_runtime": 27.3925, |
| "eval_samples_per_second": 18.253, |
| "eval_steps_per_second": 2.3, |
| "step": 3000 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 3000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 6, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.1647223387472691e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
