|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.06431281754453663, |
|
"eval_steps": 1000, |
|
"global_step": 1000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 58.440006256103516, |
|
"learning_rate": 1e-08, |
|
"loss": 1.3813, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 59.934757232666016, |
|
"learning_rate": 2.2e-08, |
|
"loss": 1.469, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 31.011018753051758, |
|
"learning_rate": 3.4500000000000005e-08, |
|
"loss": 1.2226, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 59.818233489990234, |
|
"learning_rate": 4.7e-08, |
|
"loss": 1.2458, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 60.51572036743164, |
|
"learning_rate": 5.95e-08, |
|
"loss": 1.2781, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 51.360103607177734, |
|
"learning_rate": 7.2e-08, |
|
"loss": 1.4055, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 73.35002136230469, |
|
"learning_rate": 8.45e-08, |
|
"loss": 1.3354, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 69.32823944091797, |
|
"learning_rate": 9.7e-08, |
|
"loss": 1.2005, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 51.02174377441406, |
|
"learning_rate": 1.095e-07, |
|
"loss": 1.3853, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 72.20179748535156, |
|
"learning_rate": 1.2199999999999998e-07, |
|
"loss": 1.4476, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 108.30382537841797, |
|
"learning_rate": 1.345e-07, |
|
"loss": 1.2339, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 66.15994262695312, |
|
"learning_rate": 1.4699999999999998e-07, |
|
"loss": 1.379, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 47.82923126220703, |
|
"learning_rate": 1.595e-07, |
|
"loss": 1.1467, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 85.7218246459961, |
|
"learning_rate": 1.7199999999999998e-07, |
|
"loss": 1.1622, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 68.25504302978516, |
|
"learning_rate": 1.845e-07, |
|
"loss": 1.1413, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 106.06077575683594, |
|
"learning_rate": 1.97e-07, |
|
"loss": 1.0855, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 79.60690307617188, |
|
"learning_rate": 2.095e-07, |
|
"loss": 0.929, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 42.14814376831055, |
|
"learning_rate": 2.22e-07, |
|
"loss": 0.8728, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 37.4913444519043, |
|
"learning_rate": 2.3449999999999996e-07, |
|
"loss": 0.6651, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 41.89991760253906, |
|
"learning_rate": 2.47e-07, |
|
"loss": 0.5875, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 75.21453094482422, |
|
"learning_rate": 2.595e-07, |
|
"loss": 0.6868, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 21.09180450439453, |
|
"learning_rate": 2.72e-07, |
|
"loss": 0.741, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 44.54707336425781, |
|
"learning_rate": 2.845e-07, |
|
"loss": 0.3898, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 31.656843185424805, |
|
"learning_rate": 2.9699999999999997e-07, |
|
"loss": 0.422, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 56.28642654418945, |
|
"learning_rate": 3.0949999999999996e-07, |
|
"loss": 0.3803, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 38.66410827636719, |
|
"learning_rate": 3.22e-07, |
|
"loss": 0.5062, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 31.183727264404297, |
|
"learning_rate": 3.345e-07, |
|
"loss": 0.4075, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 23.618703842163086, |
|
"learning_rate": 3.4699999999999997e-07, |
|
"loss": 0.3627, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 70.09487915039062, |
|
"learning_rate": 3.5949999999999996e-07, |
|
"loss": 0.3087, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 74.42188262939453, |
|
"learning_rate": 3.72e-07, |
|
"loss": 0.4021, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 44.99939727783203, |
|
"learning_rate": 3.845e-07, |
|
"loss": 0.3203, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 42.77998352050781, |
|
"learning_rate": 3.97e-07, |
|
"loss": 0.3797, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 64.61412811279297, |
|
"learning_rate": 4.0949999999999995e-07, |
|
"loss": 0.3403, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 29.286806106567383, |
|
"learning_rate": 4.2199999999999994e-07, |
|
"loss": 0.2879, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 58.146263122558594, |
|
"learning_rate": 4.345e-07, |
|
"loss": 0.4017, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 44.624202728271484, |
|
"learning_rate": 4.4699999999999997e-07, |
|
"loss": 0.3698, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 47.91656494140625, |
|
"learning_rate": 4.595e-07, |
|
"loss": 0.4008, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 36.263668060302734, |
|
"learning_rate": 4.7199999999999994e-07, |
|
"loss": 0.2041, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 12.398943901062012, |
|
"learning_rate": 4.845e-07, |
|
"loss": 0.2978, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 4.42283821105957, |
|
"learning_rate": 4.97e-07, |
|
"loss": 0.2614, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"eval_loss": 0.29864633083343506, |
|
"eval_runtime": 7667.7674, |
|
"eval_samples_per_second": 1.228, |
|
"eval_steps_per_second": 0.614, |
|
"eval_wer": 0.14664944291942517, |
|
"step": 1000 |
|
} |
|
], |
|
"logging_steps": 25, |
|
"max_steps": 5000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 1, |
|
"save_steps": 1000, |
|
"total_flos": 2.0412098362212352e+18, |
|
"train_batch_size": 1, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|