| { | |
| "best_metric": 0.33828356862068176, | |
| "best_model_checkpoint": "/scratch/skscla001/results/mms-zeroshot-300m-genbed-m-model/checkpoint-3400", | |
| "epoch": 11.019283746556473, | |
| "eval_steps": 200, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5509641873278237, | |
| "eval_loss": 2.310696840286255, | |
| "eval_runtime": 57.4366, | |
| "eval_samples_per_second": 16.871, | |
| "eval_steps_per_second": 2.124, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1019283746556474, | |
| "eval_loss": 0.4859784245491028, | |
| "eval_runtime": 55.9045, | |
| "eval_samples_per_second": 17.333, | |
| "eval_steps_per_second": 2.182, | |
| "eval_wer": 0.638015449896638, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3774104683195592, | |
| "grad_norm": 0.8601316213607788, | |
| "learning_rate": 0.0002889620018535681, | |
| "loss": 2.8087, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "eval_loss": 0.438145250082016, | |
| "eval_runtime": 56.5281, | |
| "eval_samples_per_second": 17.142, | |
| "eval_steps_per_second": 2.158, | |
| "eval_wer": 0.5988466978565988, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.203856749311295, | |
| "eval_loss": 0.42812615633010864, | |
| "eval_runtime": 56.0727, | |
| "eval_samples_per_second": 17.281, | |
| "eval_steps_per_second": 2.176, | |
| "eval_wer": 0.5854640409095855, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7548209366391183, | |
| "grad_norm": 0.6990365982055664, | |
| "learning_rate": 0.0002750602409638554, | |
| "loss": 0.6248, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.7548209366391183, | |
| "eval_loss": 0.4056081473827362, | |
| "eval_runtime": 56.3304, | |
| "eval_samples_per_second": 17.202, | |
| "eval_steps_per_second": 2.166, | |
| "eval_wer": 0.5655532586225656, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.3057851239669422, | |
| "eval_loss": 0.401683509349823, | |
| "eval_runtime": 56.6419, | |
| "eval_samples_per_second": 17.107, | |
| "eval_steps_per_second": 2.154, | |
| "eval_wer": 0.5513001849635513, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8567493112947657, | |
| "eval_loss": 0.39044997096061707, | |
| "eval_runtime": 56.4862, | |
| "eval_samples_per_second": 17.155, | |
| "eval_steps_per_second": 2.16, | |
| "eval_wer": 0.5617451855075617, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.132231404958677, | |
| "grad_norm": 0.828954815864563, | |
| "learning_rate": 0.0002611584800741427, | |
| "loss": 0.578, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.40771349862259, | |
| "eval_loss": 0.37762871384620667, | |
| "eval_runtime": 56.693, | |
| "eval_samples_per_second": 17.092, | |
| "eval_steps_per_second": 2.152, | |
| "eval_wer": 0.5306277880535306, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.958677685950414, | |
| "eval_loss": 0.37221014499664307, | |
| "eval_runtime": 56.5211, | |
| "eval_samples_per_second": 17.144, | |
| "eval_steps_per_second": 2.158, | |
| "eval_wer": 0.5177891415515178, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.509641873278237, | |
| "grad_norm": 0.5989723801612854, | |
| "learning_rate": 0.00024725671918443, | |
| "loss": 0.5343, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.509641873278237, | |
| "eval_loss": 0.3658897578716278, | |
| "eval_runtime": 56.2923, | |
| "eval_samples_per_second": 17.214, | |
| "eval_steps_per_second": 2.167, | |
| "eval_wer": 0.5151778914155152, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.0606060606060606, | |
| "eval_loss": 0.36135414242744446, | |
| "eval_runtime": 56.2126, | |
| "eval_samples_per_second": 17.238, | |
| "eval_steps_per_second": 2.17, | |
| "eval_wer": 0.514960287237515, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.6115702479338845, | |
| "eval_loss": 0.3572969138622284, | |
| "eval_runtime": 56.7561, | |
| "eval_samples_per_second": 17.073, | |
| "eval_steps_per_second": 2.15, | |
| "eval_wer": 0.5015776302905016, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.887052341597796, | |
| "grad_norm": 0.6925222873687744, | |
| "learning_rate": 0.0002333549582947173, | |
| "loss": 0.5153, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.162534435261708, | |
| "eval_loss": 0.3624655604362488, | |
| "eval_runtime": 56.0335, | |
| "eval_samples_per_second": 17.293, | |
| "eval_steps_per_second": 2.177, | |
| "eval_wer": 0.5025568490915026, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 7.7134986225895315, | |
| "eval_loss": 0.3545249402523041, | |
| "eval_runtime": 57.0392, | |
| "eval_samples_per_second": 16.988, | |
| "eval_steps_per_second": 2.139, | |
| "eval_wer": 0.48667174409748665, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.264462809917354, | |
| "grad_norm": 1.5090163946151733, | |
| "learning_rate": 0.00021945319740500463, | |
| "loss": 0.4935, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.264462809917354, | |
| "eval_loss": 0.350554883480072, | |
| "eval_runtime": 56.6484, | |
| "eval_samples_per_second": 17.106, | |
| "eval_steps_per_second": 2.154, | |
| "eval_wer": 0.48144924382548143, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 8.81542699724518, | |
| "eval_loss": 0.3481844663619995, | |
| "eval_runtime": 56.3422, | |
| "eval_samples_per_second": 17.198, | |
| "eval_steps_per_second": 2.165, | |
| "eval_wer": 0.49222065063649223, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 9.366391184573002, | |
| "eval_loss": 0.33828356862068176, | |
| "eval_runtime": 56.3849, | |
| "eval_samples_per_second": 17.185, | |
| "eval_steps_per_second": 2.164, | |
| "eval_wer": 0.4761179414644761, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 9.641873278236915, | |
| "grad_norm": 0.5624499917030334, | |
| "learning_rate": 0.0002055514365152919, | |
| "loss": 0.4731, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 9.917355371900827, | |
| "eval_loss": 0.34193041920661926, | |
| "eval_runtime": 56.2658, | |
| "eval_samples_per_second": 17.222, | |
| "eval_steps_per_second": 2.168, | |
| "eval_wer": 0.4644761179414645, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 10.46831955922865, | |
| "eval_loss": 0.3391243517398834, | |
| "eval_runtime": 56.3644, | |
| "eval_samples_per_second": 17.192, | |
| "eval_steps_per_second": 2.164, | |
| "eval_wer": 0.4686105973234686, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 11.019283746556473, | |
| "grad_norm": 0.5695391893386841, | |
| "learning_rate": 0.00019164967562557923, | |
| "loss": 0.4619, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 11.019283746556473, | |
| "eval_loss": 0.34022417664527893, | |
| "eval_runtime": 56.9717, | |
| "eval_samples_per_second": 17.008, | |
| "eval_steps_per_second": 2.141, | |
| "eval_wer": 0.4576215863344576, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 11.019283746556473, | |
| "step": 4000, | |
| "total_flos": 6.332346090903939e+18, | |
| "train_loss": 0.8111911582946777, | |
| "train_runtime": 3985.8224, | |
| "train_samples_per_second": 21.827, | |
| "train_steps_per_second": 2.732 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10890, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.332346090903939e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |