| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.847380410022779, | |
| "eval_steps": 200, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22779043280182232, | |
| "eval_loss": 2.9656879901885986, | |
| "eval_runtime": 180.6727, | |
| "eval_samples_per_second": 38.866, | |
| "eval_steps_per_second": 4.86, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "eval_loss": 1.256484031677246, | |
| "eval_runtime": 177.1703, | |
| "eval_samples_per_second": 39.634, | |
| "eval_steps_per_second": 4.956, | |
| "eval_wer": 0.8238084481590242, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 2.394413948059082, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.3708, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "eval_loss": 0.8798568248748779, | |
| "eval_runtime": 177.6107, | |
| "eval_samples_per_second": 39.536, | |
| "eval_steps_per_second": 4.943, | |
| "eval_wer": 0.6336517891547855, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "eval_loss": 0.7572081685066223, | |
| "eval_runtime": 179.3544, | |
| "eval_samples_per_second": 39.152, | |
| "eval_steps_per_second": 4.895, | |
| "eval_wer": 0.5467651709430101, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 0.749163806438446, | |
| "learning_rate": 0.00022664999999999995, | |
| "loss": 0.9128, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "eval_loss": 0.6133933663368225, | |
| "eval_runtime": 178.9414, | |
| "eval_samples_per_second": 39.242, | |
| "eval_steps_per_second": 4.907, | |
| "eval_wer": 0.4571280511300974, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "eval_loss": 0.5878995060920715, | |
| "eval_runtime": 179.0164, | |
| "eval_samples_per_second": 39.225, | |
| "eval_steps_per_second": 4.905, | |
| "eval_wer": 0.44289719501986474, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "eval_loss": 0.6282891631126404, | |
| "eval_runtime": 178.8026, | |
| "eval_samples_per_second": 39.272, | |
| "eval_steps_per_second": 4.91, | |
| "eval_wer": 0.46545928061760056, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 2.0273807048797607, | |
| "learning_rate": 0.00015209999999999998, | |
| "loss": 0.7779, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "eval_loss": 0.6165733337402344, | |
| "eval_runtime": 178.9936, | |
| "eval_samples_per_second": 39.23, | |
| "eval_steps_per_second": 4.905, | |
| "eval_wer": 0.4566364155782033, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "eval_loss": 0.6130892634391785, | |
| "eval_runtime": 180.5144, | |
| "eval_samples_per_second": 38.9, | |
| "eval_steps_per_second": 4.864, | |
| "eval_wer": 0.46482148314487304, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 2.559767246246338, | |
| "learning_rate": 7.725e-05, | |
| "loss": 0.7465, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.6311540603637695, | |
| "eval_runtime": 180.4386, | |
| "eval_samples_per_second": 38.916, | |
| "eval_steps_per_second": 4.866, | |
| "eval_wer": 0.4721694415285879, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.5056947608200457, | |
| "eval_loss": 0.7629979252815247, | |
| "eval_runtime": 179.3715, | |
| "eval_samples_per_second": 39.148, | |
| "eval_steps_per_second": 4.895, | |
| "eval_wer": 0.4721694415285879, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "eval_loss": 0.8067795038223267, | |
| "eval_runtime": 179.4035, | |
| "eval_samples_per_second": 39.141, | |
| "eval_steps_per_second": 4.894, | |
| "eval_wer": 0.48269309982859193, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "grad_norm": 5.402973175048828, | |
| "learning_rate": 2.9999999999999997e-06, | |
| "loss": 0.8574, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "step": 2500, | |
| "total_flos": 9.363138171711945e+18, | |
| "train_loss": 1.33308916015625, | |
| "train_runtime": 4617.7601, | |
| "train_samples_per_second": 17.324, | |
| "train_steps_per_second": 0.541 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 9.363138171711945e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |