| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 1000.0, | |
| "global_step": 5495, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09099181073703366, | |
| "grad_norm": 1.2458827495574951, | |
| "learning_rate": 7.101063243757922e-06, | |
| "loss": 0.1615, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18198362147406733, | |
| "grad_norm": 1.1883313655853271, | |
| "learning_rate": 1.3026593668101051e-05, | |
| "loss": 0.1645, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.272975432211101, | |
| "grad_norm": 1.536879539489746, | |
| "learning_rate": 1.2780576790516517e-05, | |
| "loss": 0.1829, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.36396724294813465, | |
| "grad_norm": 1.4138495922088623, | |
| "learning_rate": 1.2534559912931985e-05, | |
| "loss": 0.1856, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4549590536851683, | |
| "grad_norm": 1.4161380529403687, | |
| "learning_rate": 1.2288543035347451e-05, | |
| "loss": 0.1883, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.545950864422202, | |
| "grad_norm": 0.7464113235473633, | |
| "learning_rate": 1.2042526157762917e-05, | |
| "loss": 0.1852, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 1.0659712553024292, | |
| "learning_rate": 1.1796509280178383e-05, | |
| "loss": 0.175, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7279344858962693, | |
| "grad_norm": 0.9812346696853638, | |
| "learning_rate": 1.155049240259385e-05, | |
| "loss": 0.1738, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.818926296633303, | |
| "grad_norm": 0.6952521204948425, | |
| "learning_rate": 1.1304475525009316e-05, | |
| "loss": 0.1648, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9099181073703366, | |
| "grad_norm": 1.3266983032226562, | |
| "learning_rate": 1.1058458647424782e-05, | |
| "loss": 0.1695, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9099181073703366, | |
| "eval_cer": 0.17138804288873127, | |
| "eval_loss": 0.24572527408599854, | |
| "eval_runtime": 16.2175, | |
| "eval_samples_per_second": 30.831, | |
| "eval_steps_per_second": 0.987, | |
| "eval_wer": 0.37248797056326066, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0009099181073704, | |
| "grad_norm": 0.9983859658241272, | |
| "learning_rate": 1.081244176984025e-05, | |
| "loss": 0.1606, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.091901728844404, | |
| "grad_norm": 0.9436022639274597, | |
| "learning_rate": 1.0566424892255716e-05, | |
| "loss": 0.1479, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.1828935395814377, | |
| "grad_norm": 1.3033909797668457, | |
| "learning_rate": 1.0320408014671182e-05, | |
| "loss": 0.1526, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 1.919601559638977, | |
| "learning_rate": 1.007439113708665e-05, | |
| "loss": 0.1549, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.364877161055505, | |
| "grad_norm": 2.1733407974243164, | |
| "learning_rate": 9.828374259502114e-06, | |
| "loss": 0.1538, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4558689717925386, | |
| "grad_norm": 0.9706993699073792, | |
| "learning_rate": 9.582357381917582e-06, | |
| "loss": 0.1447, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.5468607825295724, | |
| "grad_norm": 2.0929415225982666, | |
| "learning_rate": 9.33634050433305e-06, | |
| "loss": 0.1513, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.6378525932666061, | |
| "grad_norm": 1.7381020784378052, | |
| "learning_rate": 9.090323626748514e-06, | |
| "loss": 0.1512, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.7288444040036397, | |
| "grad_norm": 4.781813621520996, | |
| "learning_rate": 8.844306749163981e-06, | |
| "loss": 0.157, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.8198362147406733, | |
| "grad_norm": 1.9795459508895874, | |
| "learning_rate": 8.598289871579447e-06, | |
| "loss": 0.1502, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.8198362147406733, | |
| "eval_cer": 0.170386749551504, | |
| "eval_loss": 0.23442533612251282, | |
| "eval_runtime": 16.2106, | |
| "eval_samples_per_second": 30.844, | |
| "eval_steps_per_second": 0.987, | |
| "eval_wer": 0.36399660345315593, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 1.2681164741516113, | |
| "learning_rate": 8.352272993994913e-06, | |
| "loss": 0.1512, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.001819836214741, | |
| "grad_norm": 0.8989790081977844, | |
| "learning_rate": 8.106256116410381e-06, | |
| "loss": 0.1498, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.092811646951774, | |
| "grad_norm": 2.716948986053467, | |
| "learning_rate": 7.860239238825847e-06, | |
| "loss": 0.1379, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.183803457688808, | |
| "grad_norm": 1.0314233303070068, | |
| "learning_rate": 7.614222361241313e-06, | |
| "loss": 0.1375, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.2747952684258417, | |
| "grad_norm": 1.034842848777771, | |
| "learning_rate": 7.3682054836567806e-06, | |
| "loss": 0.1424, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.3657870791628755, | |
| "grad_norm": 1.1963162422180176, | |
| "learning_rate": 7.122188606072246e-06, | |
| "loss": 0.1371, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.4567788898999092, | |
| "grad_norm": 1.1049108505249023, | |
| "learning_rate": 6.8761717284877125e-06, | |
| "loss": 0.1346, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 1.5126423835754395, | |
| "learning_rate": 6.630154850903179e-06, | |
| "loss": 0.1312, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.6387625113739763, | |
| "grad_norm": 1.445042610168457, | |
| "learning_rate": 6.384137973318645e-06, | |
| "loss": 0.1236, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.72975432211101, | |
| "grad_norm": 0.6577618718147278, | |
| "learning_rate": 6.138121095734112e-06, | |
| "loss": 0.1287, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.72975432211101, | |
| "eval_cer": 0.16905169176853435, | |
| "eval_loss": 0.2321455329656601, | |
| "eval_runtime": 16.3024, | |
| "eval_samples_per_second": 30.67, | |
| "eval_steps_per_second": 0.981, | |
| "eval_wer": 0.36173223889046136, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.8207461328480434, | |
| "grad_norm": 0.7591854929924011, | |
| "learning_rate": 5.892104218149578e-06, | |
| "loss": 0.1398, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.911737943585077, | |
| "grad_norm": 0.9334970116615295, | |
| "learning_rate": 5.646087340565045e-06, | |
| "loss": 0.1333, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.002729754322111, | |
| "grad_norm": 1.2228902578353882, | |
| "learning_rate": 5.400070462980511e-06, | |
| "loss": 0.1339, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.0937215650591448, | |
| "grad_norm": 0.6372168064117432, | |
| "learning_rate": 5.154053585395978e-06, | |
| "loss": 0.1219, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 1.18144690990448, | |
| "learning_rate": 4.9080367078114446e-06, | |
| "loss": 0.1253, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.275705186533212, | |
| "grad_norm": 0.8614761829376221, | |
| "learning_rate": 4.6620198302269105e-06, | |
| "loss": 0.1202, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.3666969972702456, | |
| "grad_norm": 1.8885802030563354, | |
| "learning_rate": 4.4160029526423765e-06, | |
| "loss": 0.1225, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.4576888080072794, | |
| "grad_norm": 0.8055128455162048, | |
| "learning_rate": 4.169986075057844e-06, | |
| "loss": 0.1288, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.548680618744313, | |
| "grad_norm": 0.6375197172164917, | |
| "learning_rate": 3.92396919747331e-06, | |
| "loss": 0.1291, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.6396724294813465, | |
| "grad_norm": 0.9885872602462769, | |
| "learning_rate": 3.6779523198887766e-06, | |
| "loss": 0.1248, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.6396724294813465, | |
| "eval_cer": 0.16650673786974843, | |
| "eval_loss": 0.2386896312236786, | |
| "eval_runtime": 16.2576, | |
| "eval_samples_per_second": 30.755, | |
| "eval_steps_per_second": 0.984, | |
| "eval_wer": 0.3560713274837249, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.7306642402183803, | |
| "grad_norm": 1.435037612915039, | |
| "learning_rate": 3.4319354423042425e-06, | |
| "loss": 0.1191, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 1.0509116649627686, | |
| "learning_rate": 3.1859185647197094e-06, | |
| "loss": 0.1144, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.912647861692448, | |
| "grad_norm": 1.5290184020996094, | |
| "learning_rate": 2.9399016871351758e-06, | |
| "loss": 0.1244, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 4.003639672429482, | |
| "grad_norm": 1.2121403217315674, | |
| "learning_rate": 2.693884809550642e-06, | |
| "loss": 0.1161, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.094631483166515, | |
| "grad_norm": 1.786240816116333, | |
| "learning_rate": 2.447867931966109e-06, | |
| "loss": 0.1209, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.185623293903548, | |
| "grad_norm": 1.2012931108474731, | |
| "learning_rate": 2.201851054381575e-06, | |
| "loss": 0.1176, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.276615104640582, | |
| "grad_norm": 0.8693597316741943, | |
| "learning_rate": 1.9558341767970418e-06, | |
| "loss": 0.1208, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.367606915377616, | |
| "grad_norm": 1.355154037475586, | |
| "learning_rate": 1.709817299212508e-06, | |
| "loss": 0.1119, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.45859872611465, | |
| "grad_norm": 1.7125566005706787, | |
| "learning_rate": 1.4638004216279746e-06, | |
| "loss": 0.1229, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 4.549590536851683, | |
| "grad_norm": 0.8439257144927979, | |
| "learning_rate": 1.2177835440434412e-06, | |
| "loss": 0.1127, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.549590536851683, | |
| "eval_cer": 0.16671534064833743, | |
| "eval_loss": 0.23713621497154236, | |
| "eval_runtime": 16.2282, | |
| "eval_samples_per_second": 30.811, | |
| "eval_steps_per_second": 0.986, | |
| "eval_wer": 0.3492782337956411, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.640582347588717, | |
| "grad_norm": 1.417340636253357, | |
| "learning_rate": 9.717666664589076e-07, | |
| "loss": 0.1241, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 4.731574158325751, | |
| "grad_norm": 2.014954090118408, | |
| "learning_rate": 7.25749788874374e-07, | |
| "loss": 0.1165, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 4.822565969062785, | |
| "grad_norm": 1.0176433324813843, | |
| "learning_rate": 4.797329112898405e-07, | |
| "loss": 0.1195, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 4.9135577797998184, | |
| "grad_norm": 1.0092387199401855, | |
| "learning_rate": 2.3371603370530685e-07, | |
| "loss": 0.117, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 5495, | |
| "total_flos": 4.052184710714386e+19, | |
| "train_loss": 0.13981490825066467, | |
| "train_runtime": 8884.6541, | |
| "train_samples_per_second": 19.778, | |
| "train_steps_per_second": 0.618 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5495, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.052184710714386e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |