| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 758, | |
| "global_step": 11370, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6596306068601583, | |
| "grad_norm": 1.0755512714385986, | |
| "learning_rate": 0.0002982, | |
| "loss": 3.7968, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.2848176658153534, | |
| "eval_runtime": 177.1073, | |
| "eval_samples_per_second": 64.565, | |
| "eval_steps_per_second": 4.037, | |
| "eval_wer": 0.5295409253093712, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.3192612137203166, | |
| "grad_norm": 0.398578405380249, | |
| "learning_rate": 0.00028628334866605336, | |
| "loss": 0.3537, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.978891820580475, | |
| "grad_norm": 0.6923481822013855, | |
| "learning_rate": 0.00027248390064397424, | |
| "loss": 0.2547, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.19083459675312042, | |
| "eval_runtime": 177.247, | |
| "eval_samples_per_second": 64.514, | |
| "eval_steps_per_second": 4.034, | |
| "eval_wer": 0.42224847551880096, | |
| "step": 1516 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 0.2965054214000702, | |
| "learning_rate": 0.0002586844526218951, | |
| "loss": 0.1929, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.1753465086221695, | |
| "eval_runtime": 183.1101, | |
| "eval_samples_per_second": 62.449, | |
| "eval_steps_per_second": 3.905, | |
| "eval_wer": 0.39995704006014393, | |
| "step": 2274 | |
| }, | |
| { | |
| "epoch": 3.2981530343007917, | |
| "grad_norm": 0.22291332483291626, | |
| "learning_rate": 0.000244885004599816, | |
| "loss": 0.172, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.9577836411609497, | |
| "grad_norm": 0.21266356110572815, | |
| "learning_rate": 0.00023108555657773688, | |
| "loss": 0.1532, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.15584248304367065, | |
| "eval_runtime": 179.8085, | |
| "eval_samples_per_second": 63.595, | |
| "eval_steps_per_second": 3.976, | |
| "eval_wer": 0.37095908065728705, | |
| "step": 3032 | |
| }, | |
| { | |
| "epoch": 4.617414248021108, | |
| "grad_norm": 0.3851657509803772, | |
| "learning_rate": 0.00021728610855565776, | |
| "loss": 0.1297, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.15116363763809204, | |
| "eval_runtime": 178.6057, | |
| "eval_samples_per_second": 64.024, | |
| "eval_steps_per_second": 4.003, | |
| "eval_wer": 0.35356030501557295, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 5.277044854881266, | |
| "grad_norm": 0.6844151616096497, | |
| "learning_rate": 0.00020348666053357864, | |
| "loss": 0.1236, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.936675461741425, | |
| "grad_norm": 0.2345355600118637, | |
| "learning_rate": 0.00018968721251149952, | |
| "loss": 0.1167, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.15742355585098267, | |
| "eval_runtime": 183.8285, | |
| "eval_samples_per_second": 62.205, | |
| "eval_steps_per_second": 3.889, | |
| "eval_wer": 0.35143617465602217, | |
| "step": 4548 | |
| }, | |
| { | |
| "epoch": 6.596306068601583, | |
| "grad_norm": 0.18163040280342102, | |
| "learning_rate": 0.00017588776448942042, | |
| "loss": 0.101, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.1482517421245575, | |
| "eval_runtime": 181.3176, | |
| "eval_samples_per_second": 63.066, | |
| "eval_steps_per_second": 3.943, | |
| "eval_wer": 0.3374383942529147, | |
| "step": 5306 | |
| }, | |
| { | |
| "epoch": 7.255936675461742, | |
| "grad_norm": 0.3131236135959625, | |
| "learning_rate": 0.0001620883164673413, | |
| "loss": 0.0945, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 7.915567282321899, | |
| "grad_norm": 0.5193817615509033, | |
| "learning_rate": 0.00014828886844526218, | |
| "loss": 0.0859, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.14896264672279358, | |
| "eval_runtime": 180.9851, | |
| "eval_samples_per_second": 63.182, | |
| "eval_steps_per_second": 3.951, | |
| "eval_wer": 0.32986073819496653, | |
| "step": 6064 | |
| }, | |
| { | |
| "epoch": 8.575197889182059, | |
| "grad_norm": 0.21172136068344116, | |
| "learning_rate": 0.00013448942042318306, | |
| "loss": 0.0791, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.15233299136161804, | |
| "eval_runtime": 179.7038, | |
| "eval_samples_per_second": 63.632, | |
| "eval_steps_per_second": 3.979, | |
| "eval_wer": 0.32503967827778374, | |
| "step": 6822 | |
| }, | |
| { | |
| "epoch": 9.234828496042216, | |
| "grad_norm": 0.20351053774356842, | |
| "learning_rate": 0.0001207175712971481, | |
| "loss": 0.0745, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 9.894459102902374, | |
| "grad_norm": 0.1984056681394577, | |
| "learning_rate": 0.00010694572217111315, | |
| "loss": 0.0702, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.16084744036197662, | |
| "eval_runtime": 179.8551, | |
| "eval_samples_per_second": 63.579, | |
| "eval_steps_per_second": 3.975, | |
| "eval_wer": 0.3191923531307056, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 10.554089709762533, | |
| "grad_norm": 0.9130335450172424, | |
| "learning_rate": 9.314627414903402e-05, | |
| "loss": 0.0629, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.16640283167362213, | |
| "eval_runtime": 180.2008, | |
| "eval_samples_per_second": 63.457, | |
| "eval_steps_per_second": 3.968, | |
| "eval_wer": 0.31460995954605664, | |
| "step": 8338 | |
| }, | |
| { | |
| "epoch": 11.213720316622691, | |
| "grad_norm": 0.24980618059635162, | |
| "learning_rate": 7.934682612695491e-05, | |
| "loss": 0.0587, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 11.87335092348285, | |
| "grad_norm": 0.2797747850418091, | |
| "learning_rate": 6.55473781048758e-05, | |
| "loss": 0.0559, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.16405943036079407, | |
| "eval_runtime": 182.307, | |
| "eval_samples_per_second": 62.724, | |
| "eval_steps_per_second": 3.922, | |
| "eval_wer": 0.3103020322438215, | |
| "step": 9096 | |
| }, | |
| { | |
| "epoch": 12.532981530343008, | |
| "grad_norm": 0.15323172509670258, | |
| "learning_rate": 5.1775528978840844e-05, | |
| "loss": 0.0527, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.16645778715610504, | |
| "eval_runtime": 180.7018, | |
| "eval_samples_per_second": 63.281, | |
| "eval_steps_per_second": 3.957, | |
| "eval_wer": 0.3063043711738804, | |
| "step": 9854 | |
| }, | |
| { | |
| "epoch": 13.192612137203167, | |
| "grad_norm": 0.7859821319580078, | |
| "learning_rate": 3.797608095676173e-05, | |
| "loss": 0.0491, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 13.852242744063325, | |
| "grad_norm": 0.2534639537334442, | |
| "learning_rate": 2.4176632934682608e-05, | |
| "loss": 0.0468, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.16911591589450836, | |
| "eval_runtime": 181.5239, | |
| "eval_samples_per_second": 62.994, | |
| "eval_steps_per_second": 3.939, | |
| "eval_wer": 0.30112531175789686, | |
| "step": 10612 | |
| }, | |
| { | |
| "epoch": 14.511873350923484, | |
| "grad_norm": 0.21820344030857086, | |
| "learning_rate": 1.0377184912603496e-05, | |
| "loss": 0.0443, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.17480023205280304, | |
| "eval_runtime": 182.6284, | |
| "eval_samples_per_second": 62.613, | |
| "eval_steps_per_second": 3.915, | |
| "eval_wer": 0.2998007136123343, | |
| "step": 11370 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 11370, | |
| "total_flos": 1.156593069425741e+20, | |
| "train_loss": 0.27266296556149017, | |
| "train_runtime": 21246.5656, | |
| "train_samples_per_second": 34.234, | |
| "train_steps_per_second": 0.535 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 11370, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 758, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.156593069425741e+20, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |