| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 58.8235294117647, | |
| "global_step": 9000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.96, | |
| "eval_loss": 2.9419331550598145, | |
| "eval_runtime": 165.8041, | |
| "eval_samples_per_second": 11.122, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 9.774774774774775e-05, | |
| "loss": 5.732, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "eval_loss": 2.8035287857055664, | |
| "eval_runtime": 163.2568, | |
| "eval_samples_per_second": 11.295, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "eval_loss": 0.5059587955474854, | |
| "eval_runtime": 166.512, | |
| "eval_samples_per_second": 11.074, | |
| "eval_wer": 0.646839455528622, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 9.211711711711712e-05, | |
| "loss": 1.7477, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "eval_loss": 0.31517693400382996, | |
| "eval_runtime": 168.535, | |
| "eval_samples_per_second": 10.941, | |
| "eval_wer": 0.49533842998321836, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 8.64864864864865e-05, | |
| "loss": 0.4124, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "eval_loss": 0.30525174736976624, | |
| "eval_runtime": 170.5049, | |
| "eval_samples_per_second": 10.815, | |
| "eval_wer": 0.4454596308036547, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.76, | |
| "eval_loss": 0.27497437596321106, | |
| "eval_runtime": 171.8183, | |
| "eval_samples_per_second": 10.732, | |
| "eval_wer": 0.4267201193361924, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 13.07, | |
| "learning_rate": 8.085585585585586e-05, | |
| "loss": 0.2831, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.73, | |
| "eval_loss": 0.28638383746147156, | |
| "eval_runtime": 174.9018, | |
| "eval_samples_per_second": 10.543, | |
| "eval_wer": 0.4133880290881969, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 15.69, | |
| "eval_loss": 0.27328914403915405, | |
| "eval_runtime": 174.9487, | |
| "eval_samples_per_second": 10.54, | |
| "eval_wer": 0.40583628566101065, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.34, | |
| "learning_rate": 7.522522522522523e-05, | |
| "loss": 0.2231, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.65, | |
| "eval_loss": 0.29071566462516785, | |
| "eval_runtime": 177.3068, | |
| "eval_samples_per_second": 10.4, | |
| "eval_wer": 0.3988439306358382, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 19.61, | |
| "learning_rate": 6.95945945945946e-05, | |
| "loss": 0.1885, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 19.61, | |
| "eval_loss": 0.2765745520591736, | |
| "eval_runtime": 177.2476, | |
| "eval_samples_per_second": 10.404, | |
| "eval_wer": 0.39912362483684505, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 21.57, | |
| "eval_loss": 0.28941160440444946, | |
| "eval_runtime": 180.7189, | |
| "eval_samples_per_second": 10.204, | |
| "eval_wer": 0.38355398098079435, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 22.88, | |
| "learning_rate": 6.396396396396397e-05, | |
| "loss": 0.1657, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 23.53, | |
| "eval_loss": 0.27626267075538635, | |
| "eval_runtime": 182.3178, | |
| "eval_samples_per_second": 10.114, | |
| "eval_wer": 0.38094350177139663, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 25.49, | |
| "eval_loss": 0.28250962495803833, | |
| "eval_runtime": 183.703, | |
| "eval_samples_per_second": 10.038, | |
| "eval_wer": 0.3820622785754242, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 26.14, | |
| "learning_rate": 5.833333333333334e-05, | |
| "loss": 0.1449, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.45, | |
| "eval_loss": 0.27114051580429077, | |
| "eval_runtime": 183.7677, | |
| "eval_samples_per_second": 10.034, | |
| "eval_wer": 0.38140965877307476, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 29.41, | |
| "learning_rate": 5.27027027027027e-05, | |
| "loss": 0.1344, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 29.41, | |
| "eval_loss": 0.3000315725803375, | |
| "eval_runtime": 185.7074, | |
| "eval_samples_per_second": 9.93, | |
| "eval_wer": 0.374230840947231, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 31.37, | |
| "eval_loss": 0.2914937734603882, | |
| "eval_runtime": 169.8087, | |
| "eval_samples_per_second": 10.859, | |
| "eval_wer": 0.3726459071415253, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 32.68, | |
| "learning_rate": 4.707207207207208e-05, | |
| "loss": 0.1229, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 0.2908037304878235, | |
| "eval_runtime": 189.3466, | |
| "eval_samples_per_second": 9.739, | |
| "eval_wer": 0.3699421965317919, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 35.29, | |
| "eval_loss": 0.3009437620639801, | |
| "eval_runtime": 170.798, | |
| "eval_samples_per_second": 10.796, | |
| "eval_wer": 0.3701286593324632, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 35.95, | |
| "learning_rate": 4.1441441441441444e-05, | |
| "loss": 0.1133, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 37.25, | |
| "eval_loss": 0.3189816176891327, | |
| "eval_runtime": 171.2661, | |
| "eval_samples_per_second": 10.767, | |
| "eval_wer": 0.36947603953011376, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 39.22, | |
| "learning_rate": 3.581081081081081e-05, | |
| "loss": 0.11, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 39.22, | |
| "eval_loss": 0.30211004614830017, | |
| "eval_runtime": 190.742, | |
| "eval_samples_per_second": 9.668, | |
| "eval_wer": 0.36267014730561253, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 41.18, | |
| "eval_loss": 0.30648261308670044, | |
| "eval_runtime": 173.8977, | |
| "eval_samples_per_second": 10.604, | |
| "eval_wer": 0.36677232892038036, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 42.48, | |
| "learning_rate": 3.0180180180180183e-05, | |
| "loss": 0.1006, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 43.14, | |
| "eval_loss": 0.30688807368278503, | |
| "eval_runtime": 195.1525, | |
| "eval_samples_per_second": 9.449, | |
| "eval_wer": 0.36826403132575053, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 45.1, | |
| "eval_loss": 0.31185880303382874, | |
| "eval_runtime": 176.2725, | |
| "eval_samples_per_second": 10.461, | |
| "eval_wer": 0.36164460190192055, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 45.75, | |
| "learning_rate": 2.454954954954955e-05, | |
| "loss": 0.0998, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 47.06, | |
| "eval_loss": 0.31436818838119507, | |
| "eval_runtime": 195.8727, | |
| "eval_samples_per_second": 9.414, | |
| "eval_wer": 0.36108521349990674, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 0.0952, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "eval_loss": 0.31186020374298096, | |
| "eval_runtime": 201.6879, | |
| "eval_samples_per_second": 9.143, | |
| "eval_wer": 0.3613649077009137, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 50.98, | |
| "eval_loss": 0.3063485026359558, | |
| "eval_runtime": 177.6307, | |
| "eval_samples_per_second": 10.381, | |
| "eval_wer": 0.35959351109453663, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 52.29, | |
| "learning_rate": 1.3288288288288289e-05, | |
| "loss": 0.0898, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 52.94, | |
| "eval_loss": 0.32244834303855896, | |
| "eval_runtime": 179.3678, | |
| "eval_samples_per_second": 10.281, | |
| "eval_wer": 0.360525825097893, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 54.9, | |
| "eval_loss": 0.32377830147743225, | |
| "eval_runtime": 179.7421, | |
| "eval_samples_per_second": 10.259, | |
| "eval_wer": 0.360525825097893, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 55.56, | |
| "learning_rate": 7.657657657657658e-06, | |
| "loss": 0.0891, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 56.86, | |
| "eval_loss": 0.3199419677257538, | |
| "eval_runtime": 178.805, | |
| "eval_samples_per_second": 10.313, | |
| "eval_wer": 0.3609919820995711, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 58.82, | |
| "learning_rate": 2.0270270270270273e-06, | |
| "loss": 0.0877, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 58.82, | |
| "eval_loss": 0.3155083954334259, | |
| "eval_runtime": 180.9925, | |
| "eval_samples_per_second": 10.188, | |
| "eval_wer": 0.35922058549319413, | |
| "step": 9000 | |
| } | |
| ], | |
| "max_steps": 9180, | |
| "num_train_epochs": 60, | |
| "total_flos": 6.238876246016606e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |