| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 11.999599519423308, | |
| "global_step": 7488, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.836092715231788e-05, | |
| "loss": 7.1097, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_cer": 0.9965361196010141, | |
| "eval_loss": 3.3989405632019043, | |
| "eval_runtime": 296.7149, | |
| "eval_samples_per_second": 12.952, | |
| "eval_steps_per_second": 1.621, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 9.339403973509933e-05, | |
| "loss": 3.0235, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_cer": 0.31626033596322883, | |
| "eval_loss": 1.3183486461639404, | |
| "eval_runtime": 296.8438, | |
| "eval_samples_per_second": 12.946, | |
| "eval_steps_per_second": 1.62, | |
| "eval_wer": 0.7976831955321887, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 8.84271523178808e-05, | |
| "loss": 1.1419, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_cer": 0.19131423697490937, | |
| "eval_loss": 0.6415870785713196, | |
| "eval_runtime": 297.4006, | |
| "eval_samples_per_second": 12.922, | |
| "eval_steps_per_second": 1.617, | |
| "eval_wer": 0.5543314644734832, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 8.346026490066225e-05, | |
| "loss": 0.8242, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_cer": 0.16081218755182788, | |
| "eval_loss": 0.5063228011131287, | |
| "eval_runtime": 296.5772, | |
| "eval_samples_per_second": 12.958, | |
| "eval_steps_per_second": 1.622, | |
| "eval_wer": 0.4804206025436855, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 7.683774834437087e-05, | |
| "loss": 0.6876, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_cer": 0.1387068495747151, | |
| "eval_loss": 0.44008567929267883, | |
| "eval_runtime": 296.3457, | |
| "eval_samples_per_second": 12.968, | |
| "eval_steps_per_second": 1.623, | |
| "eval_wer": 0.42799799297541397, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 7.021523178807947e-05, | |
| "loss": 0.5868, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "eval_cer": 0.12487975928163575, | |
| "eval_loss": 0.39397352933883667, | |
| "eval_runtime": 295.9984, | |
| "eval_samples_per_second": 12.983, | |
| "eval_steps_per_second": 1.625, | |
| "eval_wer": 0.3906716987717882, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 6.359271523178809e-05, | |
| "loss": 0.5285, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "eval_cer": 0.1200464377946786, | |
| "eval_loss": 0.36609867215156555, | |
| "eval_runtime": 296.3002, | |
| "eval_samples_per_second": 12.97, | |
| "eval_steps_per_second": 1.623, | |
| "eval_wer": 0.3762952943999651, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 6.433863885839737e-05, | |
| "loss": 0.5, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "eval_cer": 0.11358305494349279, | |
| "eval_loss": 0.35275039076805115, | |
| "eval_runtime": 294.0242, | |
| "eval_samples_per_second": 13.07, | |
| "eval_steps_per_second": 1.636, | |
| "eval_wer": 0.36098082418900934, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 5.885016465422613e-05, | |
| "loss": 0.4538, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "eval_cer": 0.10862653114412302, | |
| "eval_loss": 0.34030598402023315, | |
| "eval_runtime": 295.0902, | |
| "eval_samples_per_second": 13.023, | |
| "eval_steps_per_second": 1.63, | |
| "eval_wer": 0.34854599794934443, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 5.336169045005489e-05, | |
| "loss": 0.4165, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "eval_cer": 0.1061861776482574, | |
| "eval_loss": 0.33348363637924194, | |
| "eval_runtime": 294.4649, | |
| "eval_samples_per_second": 13.051, | |
| "eval_steps_per_second": 1.633, | |
| "eval_wer": 0.3438774842383124, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 4.787321624588365e-05, | |
| "loss": 0.3989, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "eval_cer": 0.1035941905371147, | |
| "eval_loss": 0.3264155387878418, | |
| "eval_runtime": 294.7693, | |
| "eval_samples_per_second": 13.037, | |
| "eval_steps_per_second": 1.632, | |
| "eval_wer": 0.3339950697004734, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 4.2384742041712404e-05, | |
| "loss": 0.3679, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "eval_cer": 0.10127703935366171, | |
| "eval_loss": 0.3256165385246277, | |
| "eval_runtime": 293.9503, | |
| "eval_samples_per_second": 13.074, | |
| "eval_steps_per_second": 1.636, | |
| "eval_wer": 0.3287375379044045, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 3.689626783754117e-05, | |
| "loss": 0.3517, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_cer": 0.10020612694576729, | |
| "eval_loss": 0.3212122619152069, | |
| "eval_runtime": 293.3063, | |
| "eval_samples_per_second": 13.102, | |
| "eval_steps_per_second": 1.64, | |
| "eval_wer": 0.32230196993826216, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 3.142151481888035e-05, | |
| "loss": 0.3357, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "eval_cer": 0.09861871253583529, | |
| "eval_loss": 0.317330539226532, | |
| "eval_runtime": 293.5785, | |
| "eval_samples_per_second": 13.09, | |
| "eval_steps_per_second": 1.638, | |
| "eval_wer": 0.3196404808132813, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "learning_rate": 2.5933040614709113e-05, | |
| "loss": 0.3225, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_cer": 0.09852868007676452, | |
| "eval_loss": 0.31416377425193787, | |
| "eval_runtime": 294.148, | |
| "eval_samples_per_second": 13.065, | |
| "eval_steps_per_second": 1.635, | |
| "eval_wer": 0.31769890268112305, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 2.044456641053787e-05, | |
| "loss": 0.3057, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "eval_cer": 0.09754780012794086, | |
| "eval_loss": 0.31993794441223145, | |
| "eval_runtime": 293.1881, | |
| "eval_samples_per_second": 13.108, | |
| "eval_steps_per_second": 1.641, | |
| "eval_wer": 0.3155828006719169, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "learning_rate": 1.4956092206366632e-05, | |
| "loss": 0.2972, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "eval_cer": 0.09668064539057503, | |
| "eval_loss": 0.3139244616031647, | |
| "eval_runtime": 292.3994, | |
| "eval_samples_per_second": 13.143, | |
| "eval_steps_per_second": 1.645, | |
| "eval_wer": 0.3127686031545191, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "learning_rate": 9.46761800219539e-06, | |
| "loss": 0.2881, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "eval_cer": 0.09570450399222878, | |
| "eval_loss": 0.31835824251174927, | |
| "eval_runtime": 293.4248, | |
| "eval_samples_per_second": 13.097, | |
| "eval_steps_per_second": 1.639, | |
| "eval_wer": 0.31065250114531295, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 3.979143798024149e-06, | |
| "loss": 0.2791, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "eval_cer": 0.09584192195607363, | |
| "eval_loss": 0.3184186816215515, | |
| "eval_runtime": 293.2834, | |
| "eval_samples_per_second": 13.103, | |
| "eval_steps_per_second": 1.64, | |
| "eval_wer": 0.31036889984511007, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 7488, | |
| "total_flos": 5.018412278085267e+19, | |
| "train_loss": 0.24150397533025497, | |
| "train_runtime": 25397.5173, | |
| "train_samples_per_second": 18.876, | |
| "train_steps_per_second": 0.295 | |
| } | |
| ], | |
| "max_steps": 7488, | |
| "num_train_epochs": 12, | |
| "total_flos": 5.018412278085267e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |