| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.1422044545973729, | |
| "eval_steps": 500, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0713877784123358, | |
| "grad_norm": 5.507429599761963, | |
| "learning_rate": 0.00018487499999999998, | |
| "loss": 4.6973, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0713877784123358, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 112.8446, | |
| "eval_samples_per_second": 34.57, | |
| "eval_steps_per_second": 0.541, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1427755568246716, | |
| "grad_norm": 17.29350471496582, | |
| "learning_rate": 0.0002919583333333333, | |
| "loss": 1.448, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1427755568246716, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 111.1281, | |
| "eval_samples_per_second": 35.104, | |
| "eval_steps_per_second": 0.549, | |
| "eval_wer": 0.7573781577549148, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.21416333523700742, | |
| "grad_norm": 6.642392158508301, | |
| "learning_rate": 0.000271125, | |
| "loss": 1.053, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21416333523700742, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 112.2463, | |
| "eval_samples_per_second": 34.754, | |
| "eval_steps_per_second": 0.543, | |
| "eval_wer": 0.6583828430662223, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2855511136493432, | |
| "grad_norm": 7.0610857009887695, | |
| "learning_rate": 0.0002502916666666666, | |
| "loss": 0.9304, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2855511136493432, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 112.8186, | |
| "eval_samples_per_second": 34.578, | |
| "eval_steps_per_second": 0.541, | |
| "eval_wer": 0.5963386948751389, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.35693889206167906, | |
| "grad_norm": 8.276483535766602, | |
| "learning_rate": 0.0002295, | |
| "loss": 0.8755, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.35693889206167906, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 111.0309, | |
| "eval_samples_per_second": 35.134, | |
| "eval_steps_per_second": 0.549, | |
| "eval_wer": 0.5945515142732938, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.42832667047401485, | |
| "grad_norm": 7.304489612579346, | |
| "learning_rate": 0.00020874999999999998, | |
| "loss": 0.8238, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.42832667047401485, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.1934, | |
| "eval_samples_per_second": 36.392, | |
| "eval_steps_per_second": 0.569, | |
| "eval_wer": 0.5392213688837366, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.49971444888635064, | |
| "grad_norm": 7.1927947998046875, | |
| "learning_rate": 0.00018795833333333333, | |
| "loss": 0.7819, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.49971444888635064, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.346, | |
| "eval_samples_per_second": 36.34, | |
| "eval_steps_per_second": 0.568, | |
| "eval_wer": 0.49666714968845094, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5711022272986864, | |
| "grad_norm": 4.308356761932373, | |
| "learning_rate": 0.000167125, | |
| "loss": 0.729, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5711022272986864, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.5572, | |
| "eval_samples_per_second": 36.269, | |
| "eval_steps_per_second": 0.567, | |
| "eval_wer": 0.4834082017098971, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.6424900057110223, | |
| "grad_norm": 6.571809768676758, | |
| "learning_rate": 0.0001463333333333333, | |
| "loss": 0.6923, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6424900057110223, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.6399, | |
| "eval_samples_per_second": 36.241, | |
| "eval_steps_per_second": 0.567, | |
| "eval_wer": 0.4564314350577211, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.7138777841233581, | |
| "grad_norm": 5.07558536529541, | |
| "learning_rate": 0.0001255, | |
| "loss": 0.7052, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7138777841233581, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.7831, | |
| "eval_samples_per_second": 36.193, | |
| "eval_steps_per_second": 0.566, | |
| "eval_wer": 0.43462300149736754, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7852655625356939, | |
| "grad_norm": 4.797505855560303, | |
| "learning_rate": 0.00010466666666666667, | |
| "loss": 0.6675, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7852655625356939, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 108.5624, | |
| "eval_samples_per_second": 35.933, | |
| "eval_steps_per_second": 0.562, | |
| "eval_wer": 0.41626817369463365, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.8566533409480297, | |
| "grad_norm": 4.757002830505371, | |
| "learning_rate": 8.383333333333333e-05, | |
| "loss": 0.6217, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8566533409480297, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.7978, | |
| "eval_samples_per_second": 36.188, | |
| "eval_steps_per_second": 0.566, | |
| "eval_wer": 0.39619861855769695, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.9280411193603655, | |
| "grad_norm": 3.95064640045166, | |
| "learning_rate": 6.299999999999999e-05, | |
| "loss": 0.5954, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.9280411193603655, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 108.005, | |
| "eval_samples_per_second": 36.119, | |
| "eval_steps_per_second": 0.565, | |
| "eval_wer": 0.3882529102062503, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.9994288977727013, | |
| "grad_norm": 5.451465129852295, | |
| "learning_rate": 4.216666666666666e-05, | |
| "loss": 0.5687, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9994288977727013, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 108.8578, | |
| "eval_samples_per_second": 35.836, | |
| "eval_steps_per_second": 0.56, | |
| "eval_wer": 0.3746075448002705, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.070816676185037, | |
| "grad_norm": 2.5018062591552734, | |
| "learning_rate": 2.133333333333333e-05, | |
| "loss": 0.477, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.070816676185037, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 106.7519, | |
| "eval_samples_per_second": 36.543, | |
| "eval_steps_per_second": 0.571, | |
| "eval_wer": 0.36472974931169394, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.1422044545973729, | |
| "grad_norm": 4.154621124267578, | |
| "learning_rate": 5e-07, | |
| "loss": 0.4804, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1422044545973729, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.9663, | |
| "eval_samples_per_second": 36.132, | |
| "eval_steps_per_second": 0.565, | |
| "eval_wer": 0.3598029271120127, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1422044545973729, | |
| "step": 8000, | |
| "total_flos": 8.368100900404746e+18, | |
| "train_loss": 1.0091876525878907, | |
| "train_runtime": 5013.5506, | |
| "train_samples_per_second": 12.765, | |
| "train_steps_per_second": 1.596 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 8.368100900404746e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |