| { | |
| "best_metric": 0.3019354045391083, | |
| "best_model_checkpoint": "./xls-r-1b-bem-sv-female/checkpoint-2000", | |
| "epoch": 4.995196926032661, | |
| "eval_steps": 500, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.38424591738712777, | |
| "grad_norm": 4.829944610595703, | |
| "learning_rate": 1.97e-05, | |
| "loss": 4.4978, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7684918347742555, | |
| "grad_norm": 6.936720848083496, | |
| "learning_rate": 3.97e-05, | |
| "loss": 0.6184, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9606147934678194, | |
| "eval_loss": 0.5192012786865234, | |
| "eval_runtime": 63.7121, | |
| "eval_samples_per_second": 7.832, | |
| "eval_steps_per_second": 1.962, | |
| "eval_wer": 0.9458917835671342, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1527377521613833, | |
| "grad_norm": 1.6504590511322021, | |
| "learning_rate": 4.769047619047619e-05, | |
| "loss": 0.2678, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.536983669548511, | |
| "grad_norm": 0.48046204447746277, | |
| "learning_rate": 4.292857142857143e-05, | |
| "loss": 0.2085, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9212295869356388, | |
| "grad_norm": 1.0426075458526611, | |
| "learning_rate": 3.816666666666667e-05, | |
| "loss": 0.1841, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9212295869356388, | |
| "eval_loss": 0.33747053146362305, | |
| "eval_runtime": 61.1182, | |
| "eval_samples_per_second": 8.165, | |
| "eval_steps_per_second": 2.045, | |
| "eval_wer": 0.8396793587174348, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.3054755043227666, | |
| "grad_norm": 0.9610881209373474, | |
| "learning_rate": 3.3404761904761904e-05, | |
| "loss": 0.141, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.689721421709894, | |
| "grad_norm": 0.4008996784687042, | |
| "learning_rate": 2.8642857142857144e-05, | |
| "loss": 0.1244, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.881844380403458, | |
| "eval_loss": 0.31584659218788147, | |
| "eval_runtime": 61.6779, | |
| "eval_samples_per_second": 8.09, | |
| "eval_steps_per_second": 2.027, | |
| "eval_wer": 0.8396793587174348, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.073967339097022, | |
| "grad_norm": 0.6827730536460876, | |
| "learning_rate": 2.3880952380952383e-05, | |
| "loss": 0.1118, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.4582132564841497, | |
| "grad_norm": 0.3247419595718384, | |
| "learning_rate": 1.911904761904762e-05, | |
| "loss": 0.079, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.8424591738712777, | |
| "grad_norm": 0.5847994089126587, | |
| "learning_rate": 1.4357142857142858e-05, | |
| "loss": 0.0757, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.8424591738712777, | |
| "eval_loss": 0.3019354045391083, | |
| "eval_runtime": 61.5577, | |
| "eval_samples_per_second": 8.106, | |
| "eval_steps_per_second": 2.031, | |
| "eval_wer": 0.8036072144288577, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.226705091258405, | |
| "grad_norm": 0.3855469822883606, | |
| "learning_rate": 9.595238095238096e-06, | |
| "loss": 0.0553, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.610951008645533, | |
| "grad_norm": 0.22909563779830933, | |
| "learning_rate": 4.833333333333333e-06, | |
| "loss": 0.0419, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.803073967339097, | |
| "eval_loss": 0.33066290616989136, | |
| "eval_runtime": 61.6863, | |
| "eval_samples_per_second": 8.089, | |
| "eval_steps_per_second": 2.026, | |
| "eval_wer": 0.7995991983967936, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.995196926032661, | |
| "grad_norm": 0.05295922979712486, | |
| "learning_rate": 7.142857142857144e-08, | |
| "loss": 0.0363, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.995196926032661, | |
| "step": 2600, | |
| "total_flos": 1.1893077312531345e+19, | |
| "train_loss": 0.49554201878034154, | |
| "train_runtime": 6582.5135, | |
| "train_samples_per_second": 3.162, | |
| "train_steps_per_second": 0.395 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1893077312531345e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |