| { | |
| "best_metric": 3.3327760696411133, | |
| "best_model_checkpoint": "nrshoudi/hubert-xlarge-ll60k_arabic/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": 4.987782529016494e-05, | |
| "loss": 13.7253, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 3.3327760696411133, | |
| "eval_per": 1.0, | |
| "eval_runtime": 121.8139, | |
| "eval_samples_per_second": 6.674, | |
| "eval_steps_per_second": 3.341, | |
| "eval_wer": 1.0, | |
| "step": 1637 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 9.987782529016494e-05, | |
| "loss": 3.3354, | |
| "step": 3274 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 3.2846710681915283, | |
| "eval_per": 1.0, | |
| "eval_runtime": 123.7338, | |
| "eval_samples_per_second": 6.571, | |
| "eval_steps_per_second": 3.289, | |
| "eval_wer": 1.0, | |
| "step": 3274 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 9.44580194122039e-05, | |
| "loss": 3.304, | |
| "step": 4911 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 3.2374837398529053, | |
| "eval_per": 1.0, | |
| "eval_runtime": 124.396, | |
| "eval_samples_per_second": 6.536, | |
| "eval_steps_per_second": 3.272, | |
| "eval_wer": 1.0, | |
| "step": 4911 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 8.89058575985882e-05, | |
| "loss": 3.2655, | |
| "step": 6548 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.2142696380615234, | |
| "eval_per": 1.0, | |
| "eval_runtime": 124.9199, | |
| "eval_samples_per_second": 6.508, | |
| "eval_steps_per_second": 3.258, | |
| "eval_wer": 1.0, | |
| "step": 6548 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 8.335030204303264e-05, | |
| "loss": 3.2242, | |
| "step": 8185 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 3.1873738765716553, | |
| "eval_per": 1.0, | |
| "eval_runtime": 125.5279, | |
| "eval_samples_per_second": 6.477, | |
| "eval_steps_per_second": 3.242, | |
| "eval_wer": 1.0, | |
| "step": 8185 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 7.779814022941696e-05, | |
| "loss": 3.1556, | |
| "step": 9822 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 3.0734314918518066, | |
| "eval_per": 1.0, | |
| "eval_runtime": 126.5196, | |
| "eval_samples_per_second": 6.426, | |
| "eval_steps_per_second": 3.217, | |
| "eval_wer": 1.0, | |
| "step": 9822 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 7.22425846738614e-05, | |
| "loss": 3.0485, | |
| "step": 11459 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.9547784328460693, | |
| "eval_per": 1.0, | |
| "eval_runtime": 127.0062, | |
| "eval_samples_per_second": 6.401, | |
| "eval_steps_per_second": 3.205, | |
| "eval_wer": 1.0, | |
| "step": 11459 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 6.668702911830585e-05, | |
| "loss": 2.935, | |
| "step": 13096 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.83778977394104, | |
| "eval_per": 0.9171063789691888, | |
| "eval_runtime": 127.5426, | |
| "eval_samples_per_second": 6.374, | |
| "eval_steps_per_second": 3.191, | |
| "eval_wer": 0.9043441743269096, | |
| "step": 13096 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 6.113826104663001e-05, | |
| "loss": 2.8158, | |
| "step": 14733 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.702333450317383, | |
| "eval_per": 0.9086733251672477, | |
| "eval_runtime": 127.7194, | |
| "eval_samples_per_second": 6.366, | |
| "eval_steps_per_second": 3.187, | |
| "eval_wer": 0.8971116606784565, | |
| "step": 14733 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5.558270549107446e-05, | |
| "loss": 2.7185, | |
| "step": 16370 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.598202705383301, | |
| "eval_per": 0.8996749269763498, | |
| "eval_runtime": 127.815, | |
| "eval_samples_per_second": 6.361, | |
| "eval_steps_per_second": 3.184, | |
| "eval_wer": 0.8864728664084737, | |
| "step": 16370 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 5.00271499355189e-05, | |
| "loss": 2.6406, | |
| "step": 18007 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 2.5032498836517334, | |
| "eval_per": 0.857085649674927, | |
| "eval_runtime": 128.216, | |
| "eval_samples_per_second": 6.341, | |
| "eval_steps_per_second": 3.174, | |
| "eval_wer": 0.8452708693014792, | |
| "step": 18007 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 4.4478381863843074e-05, | |
| "loss": 2.5721, | |
| "step": 19644 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 2.4384822845458984, | |
| "eval_per": 0.8291011024215584, | |
| "eval_runtime": 128.477, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 3.168, | |
| "eval_wer": 0.8179273015724885, | |
| "step": 19644 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 3.892622005022738e-05, | |
| "loss": 2.5136, | |
| "step": 21281 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 2.381009578704834, | |
| "eval_per": 0.8097145010835768, | |
| "eval_runtime": 128.3969, | |
| "eval_samples_per_second": 6.332, | |
| "eval_steps_per_second": 3.17, | |
| "eval_wer": 0.7993560729783957, | |
| "step": 21281 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 3.337405823661169e-05, | |
| "loss": 2.457, | |
| "step": 22918 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 2.304086208343506, | |
| "eval_per": 0.7896212192593989, | |
| "eval_runtime": 128.5082, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 3.167, | |
| "eval_wer": 0.779151696141104, | |
| "step": 22918 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 2.7818502681056137e-05, | |
| "loss": 2.4104, | |
| "step": 24555 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 2.246638536453247, | |
| "eval_per": 0.7694808254028078, | |
| "eval_runtime": 128.3201, | |
| "eval_samples_per_second": 6.336, | |
| "eval_steps_per_second": 3.172, | |
| "eval_wer": 0.7587606737903038, | |
| "step": 24555 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 2.2262947125500577e-05, | |
| "loss": 2.3711, | |
| "step": 26192 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 2.2065765857696533, | |
| "eval_per": 0.7417318383115048, | |
| "eval_runtime": 128.1923, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 3.175, | |
| "eval_wer": 0.7321170267369698, | |
| "step": 26192 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 1.6714179053824745e-05, | |
| "loss": 2.3365, | |
| "step": 27829 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 2.1757304668426514, | |
| "eval_per": 0.7071987185527183, | |
| "eval_runtime": 128.7988, | |
| "eval_samples_per_second": 6.312, | |
| "eval_steps_per_second": 3.16, | |
| "eval_wer": 0.7002006439270216, | |
| "step": 27829 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 1.1158623498269193e-05, | |
| "loss": 2.3109, | |
| "step": 29466 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 2.141193389892578, | |
| "eval_per": 0.6828182417789503, | |
| "eval_runtime": 128.9306, | |
| "eval_samples_per_second": 6.306, | |
| "eval_steps_per_second": 3.157, | |
| "eval_wer": 0.6771499230087257, | |
| "step": 29466 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 5.603067942713636e-06, | |
| "loss": 2.2857, | |
| "step": 31103 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 2.1265432834625244, | |
| "eval_per": 0.6784603787807406, | |
| "eval_runtime": 129.2068, | |
| "eval_samples_per_second": 6.292, | |
| "eval_steps_per_second": 3.15, | |
| "eval_wer": 0.6722038169007513, | |
| "step": 31103 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 5.090612909794339e-08, | |
| "loss": 2.2757, | |
| "step": 32740 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 2.113328695297241, | |
| "eval_per": 0.670616225383963, | |
| "eval_runtime": 128.9922, | |
| "eval_samples_per_second": 6.303, | |
| "eval_steps_per_second": 3.155, | |
| "eval_wer": 0.6645980122252811, | |
| "step": 32740 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 32740, | |
| "total_flos": 3.2217850292594e+19, | |
| "train_loss": 3.285075871508953, | |
| "train_runtime": 18041.7674, | |
| "train_samples_per_second": 3.628, | |
| "train_steps_per_second": 1.815 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 32740, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 3.2217850292594e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |