| { | |
| "best_metric": 3.3562777042388916, | |
| "best_model_checkpoint": "nrshoudi/hubert-large-ls960-ft/checkpoint-1637", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 32740, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.984728161270617e-05, | |
| "loss": 14.5096, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 3.3562777042388916, | |
| "eval_per": 1.0, | |
| "eval_runtime": 97.5003, | |
| "eval_samples_per_second": 8.338, | |
| "eval_steps_per_second": 4.174, | |
| "eval_wer": 1.0, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 9.984728161270619e-05, | |
| "loss": 3.3754, | |
| "step": 3274 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 3.3135287761688232, | |
| "eval_per": 1.0, | |
| "eval_runtime": 96.1335, | |
| "eval_samples_per_second": 8.457, | |
| "eval_steps_per_second": 4.234, | |
| "eval_wer": 1.0, | |
| "step": 3274 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 9.446480689608362e-05, | |
| "loss": 3.3037, | |
| "step": 4911 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 3.2654213905334473, | |
| "eval_per": 1.0, | |
| "eval_runtime": 99.0991, | |
| "eval_samples_per_second": 8.204, | |
| "eval_steps_per_second": 4.107, | |
| "eval_wer": 1.0, | |
| "step": 4911 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 8.890925134052806e-05, | |
| "loss": 3.2641, | |
| "step": 6548 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.2260236740112305, | |
| "eval_per": 1.0, | |
| "eval_runtime": 96.3924, | |
| "eval_samples_per_second": 8.434, | |
| "eval_steps_per_second": 4.222, | |
| "eval_wer": 1.0, | |
| "step": 6548 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 8.335369578497252e-05, | |
| "loss": 3.1723, | |
| "step": 8185 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 3.1157803535461426, | |
| "eval_per": 0.9980212946386507, | |
| "eval_runtime": 99.3924, | |
| "eval_samples_per_second": 8.18, | |
| "eval_steps_per_second": 4.095, | |
| "eval_wer": 0.9974802855676357, | |
| "step": 8185 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 7.779814022941696e-05, | |
| "loss": 3.0864, | |
| "step": 9822 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 3.0251047611236572, | |
| "eval_per": 0.9234429473287478, | |
| "eval_runtime": 99.1552, | |
| "eval_samples_per_second": 8.199, | |
| "eval_steps_per_second": 4.105, | |
| "eval_wer": 0.910456814894312, | |
| "step": 9822 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 7.22425846738614e-05, | |
| "loss": 3.0205, | |
| "step": 11459 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.9357450008392334, | |
| "eval_per": 0.9068359559031377, | |
| "eval_runtime": 101.5102, | |
| "eval_samples_per_second": 8.009, | |
| "eval_steps_per_second": 4.009, | |
| "eval_wer": 0.8956651579487658, | |
| "step": 11459 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 6.669042286024571e-05, | |
| "loss": 2.9474, | |
| "step": 13096 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.852294683456421, | |
| "eval_per": 0.8527748987091303, | |
| "eval_runtime": 102.2076, | |
| "eval_samples_per_second": 7.954, | |
| "eval_steps_per_second": 3.982, | |
| "eval_wer": 0.845597498950119, | |
| "step": 13096 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 6.114165478856988e-05, | |
| "loss": 2.883, | |
| "step": 14733 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.775151252746582, | |
| "eval_per": 0.8322811646094412, | |
| "eval_runtime": 100.5566, | |
| "eval_samples_per_second": 8.085, | |
| "eval_steps_per_second": 4.047, | |
| "eval_wer": 0.8260463814101069, | |
| "step": 14733 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5.558609923301432e-05, | |
| "loss": 2.8246, | |
| "step": 16370 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.7172954082489014, | |
| "eval_per": 0.8201733722792801, | |
| "eval_runtime": 99.3308, | |
| "eval_samples_per_second": 8.185, | |
| "eval_steps_per_second": 4.097, | |
| "eval_wer": 0.8138211002753021, | |
| "step": 16370 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 5.0030543677458766e-05, | |
| "loss": 2.7959, | |
| "step": 18007 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 2.687734603881836, | |
| "eval_per": 0.8139074719683407, | |
| "eval_runtime": 99.0993, | |
| "eval_samples_per_second": 8.204, | |
| "eval_steps_per_second": 4.107, | |
| "eval_wer": 0.8073818300592599, | |
| "step": 18007 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 4.448177560578294e-05, | |
| "loss": 2.7595, | |
| "step": 19644 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 2.645852565765381, | |
| "eval_per": 0.8223640817864882, | |
| "eval_runtime": 98.3259, | |
| "eval_samples_per_second": 8.268, | |
| "eval_steps_per_second": 4.139, | |
| "eval_wer": 0.8131211795996454, | |
| "step": 19644 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 3.892622005022738e-05, | |
| "loss": 2.7335, | |
| "step": 21281 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 2.6207735538482666, | |
| "eval_per": 0.8119994346556111, | |
| "eval_runtime": 99.8187, | |
| "eval_samples_per_second": 8.145, | |
| "eval_steps_per_second": 4.077, | |
| "eval_wer": 0.8049554383836498, | |
| "step": 21281 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 3.337405823661169e-05, | |
| "loss": 2.7014, | |
| "step": 22918 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 2.579977035522461, | |
| "eval_per": 0.805356638085367, | |
| "eval_runtime": 99.5509, | |
| "eval_samples_per_second": 8.167, | |
| "eval_steps_per_second": 4.088, | |
| "eval_wer": 0.7985628295459848, | |
| "step": 22918 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 2.7818502681056137e-05, | |
| "loss": 2.6792, | |
| "step": 24555 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 2.560378074645996, | |
| "eval_per": 0.8058042023932912, | |
| "eval_runtime": 98.6889, | |
| "eval_samples_per_second": 8.238, | |
| "eval_steps_per_second": 4.124, | |
| "eval_wer": 0.7989361205730017, | |
| "step": 24555 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 2.226634086744044e-05, | |
| "loss": 2.6539, | |
| "step": 26192 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 2.528850555419922, | |
| "eval_per": 0.7984782813530575, | |
| "eval_runtime": 91.0813, | |
| "eval_samples_per_second": 8.926, | |
| "eval_steps_per_second": 4.469, | |
| "eval_wer": 0.7929168027623535, | |
| "step": 26192 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 1.6714179053824745e-05, | |
| "loss": 2.6316, | |
| "step": 27829 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 2.5119011402130127, | |
| "eval_per": 0.8017525676057665, | |
| "eval_runtime": 90.1443, | |
| "eval_samples_per_second": 9.019, | |
| "eval_steps_per_second": 4.515, | |
| "eval_wer": 0.79426998273529, | |
| "step": 27829 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 1.1158623498269193e-05, | |
| "loss": 2.6212, | |
| "step": 29466 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 2.496473789215088, | |
| "eval_per": 0.8005276547630265, | |
| "eval_runtime": 91.8076, | |
| "eval_samples_per_second": 8.855, | |
| "eval_steps_per_second": 4.433, | |
| "eval_wer": 0.7931034482758621, | |
| "step": 29466 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 5.603067942713636e-06, | |
| "loss": 2.6114, | |
| "step": 31103 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 2.4890170097351074, | |
| "eval_per": 0.7959342316027513, | |
| "eval_runtime": 92.8676, | |
| "eval_samples_per_second": 8.754, | |
| "eval_steps_per_second": 4.383, | |
| "eval_wer": 0.7893705380056927, | |
| "step": 31103 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 5.090612909794339e-08, | |
| "loss": 2.6028, | |
| "step": 32740 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 2.484998941421509, | |
| "eval_per": 0.7944737585979459, | |
| "eval_runtime": 94.0732, | |
| "eval_samples_per_second": 8.642, | |
| "eval_steps_per_second": 4.326, | |
| "eval_wer": 0.7884373104381504, | |
| "step": 32740 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 32740, | |
| "total_flos": 1.0578414234417158e+19, | |
| "train_loss": 3.458883378035278, | |
| "train_runtime": 12432.4169, | |
| "train_samples_per_second": 5.265, | |
| "train_steps_per_second": 2.633 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 32740, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 1.0578414234417158e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |