| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.999599519423308, | |
| "global_step": 4992, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 9.386753883892068e-05, | |
| "loss": 0.9821, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_cer": 0.15726841629986146, | |
| "eval_loss": 0.5059255957603455, | |
| "eval_runtime": 297.9957, | |
| "eval_samples_per_second": 12.896, | |
| "eval_steps_per_second": 0.406, | |
| "eval_wer": 0.47830450053447937, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 8.569092395748161e-05, | |
| "loss": 0.6861, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_cer": 0.13556268941095262, | |
| "eval_loss": 0.42013171315193176, | |
| "eval_runtime": 295.6579, | |
| "eval_samples_per_second": 12.998, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.4247474857653963, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 7.751430907604253e-05, | |
| "loss": 0.585, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_cer": 0.1209960711263244, | |
| "eval_loss": 0.3797381818294525, | |
| "eval_runtime": 295.6141, | |
| "eval_samples_per_second": 13.0, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.3810947010187831, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 6.933769419460344e-05, | |
| "loss": 0.5193, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_cer": 0.11522804960498509, | |
| "eval_loss": 0.3577031195163727, | |
| "eval_runtime": 295.3778, | |
| "eval_samples_per_second": 13.01, | |
| "eval_steps_per_second": 0.41, | |
| "eval_wer": 0.36521302820742163, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 6.116107931316435e-05, | |
| "loss": 0.4583, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "eval_cer": 0.11111951624265405, | |
| "eval_loss": 0.34224170446395874, | |
| "eval_runtime": 295.5987, | |
| "eval_samples_per_second": 13.001, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.35192739806714807, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 5.298446443172527e-05, | |
| "loss": 0.4282, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "eval_cer": 0.10712445215615264, | |
| "eval_loss": 0.32608968019485474, | |
| "eval_runtime": 295.3154, | |
| "eval_samples_per_second": 13.013, | |
| "eval_steps_per_second": 0.41, | |
| "eval_wer": 0.34501188943912386, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 4.4807849550286185e-05, | |
| "loss": 0.3951, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "eval_cer": 0.1047794204556737, | |
| "eval_loss": 0.3201062083244324, | |
| "eval_runtime": 296.053, | |
| "eval_samples_per_second": 12.981, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.332489801260935, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 3.66312346688471e-05, | |
| "loss": 0.3619, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "eval_cer": 0.1030111909073287, | |
| "eval_loss": 0.3167116940021515, | |
| "eval_runtime": 295.6579, | |
| "eval_samples_per_second": 12.998, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.3296101572896442, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 2.8454619787408017e-05, | |
| "loss": 0.345, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "eval_cer": 0.10128551233741981, | |
| "eval_loss": 0.31573495268821716, | |
| "eval_runtime": 296.0935, | |
| "eval_samples_per_second": 12.979, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.3209712253757717, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 2.027800490596893e-05, | |
| "loss": 0.338, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "eval_cer": 0.0982123861170341, | |
| "eval_loss": 0.305124431848526, | |
| "eval_runtime": 295.7779, | |
| "eval_samples_per_second": 12.993, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.31429568707868843, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 1.2121831561733443e-05, | |
| "loss": 0.3155, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "eval_cer": 0.09860007280945199, | |
| "eval_loss": 0.30590805411338806, | |
| "eval_runtime": 296.0673, | |
| "eval_samples_per_second": 12.98, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.315408276794869, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 3.945216680294359e-06, | |
| "loss": 0.3057, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_cer": 0.09723844149949648, | |
| "eval_loss": 0.3034806251525879, | |
| "eval_runtime": 295.665, | |
| "eval_samples_per_second": 12.998, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.3137066689936517, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 4992, | |
| "total_flos": 3.347893330222721e+19, | |
| "train_loss": 0.47035410618170714, | |
| "train_runtime": 23681.5399, | |
| "train_samples_per_second": 13.496, | |
| "train_steps_per_second": 0.211 | |
| } | |
| ], | |
| "max_steps": 4992, | |
| "num_train_epochs": 8, | |
| "total_flos": 3.347893330222721e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |