| { | |
| "best_metric": 0.2248678207397461, | |
| "best_model_checkpoint": "./checkpoint-1600", | |
| "epoch": 61.53333333333333, | |
| "global_step": 1600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 4.980500000000001e-05, | |
| "loss": 4.9923, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.156188488006592, | |
| "eval_runtime": 15.7049, | |
| "eval_samples_per_second": 21.331, | |
| "eval_steps_per_second": 0.382, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 8e-05, | |
| "loss": 2.1775, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_cer": 0.11219018715225089, | |
| "eval_loss": 0.43336454033851624, | |
| "eval_runtime": 15.2789, | |
| "eval_samples_per_second": 21.926, | |
| "eval_steps_per_second": 0.393, | |
| "eval_wer": 0.5804059328649492, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 11.53, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3708, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.53, | |
| "eval_cer": 0.07966616084977238, | |
| "eval_loss": 0.3105751574039459, | |
| "eval_runtime": 15.048, | |
| "eval_samples_per_second": 22.262, | |
| "eval_steps_per_second": 0.399, | |
| "eval_wer": 0.4336455893832943, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2266, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "eval_cer": 0.06732422862923622, | |
| "eval_loss": 0.26751142740249634, | |
| "eval_runtime": 15.232, | |
| "eval_samples_per_second": 21.993, | |
| "eval_steps_per_second": 0.394, | |
| "eval_wer": 0.3672911787665886, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "learning_rate": 8e-05, | |
| "loss": 1.093, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "eval_cer": 0.06327769347496207, | |
| "eval_loss": 0.24162611365318298, | |
| "eval_runtime": 14.995, | |
| "eval_samples_per_second": 22.341, | |
| "eval_steps_per_second": 0.4, | |
| "eval_wer": 0.35011709601873536, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 8e-05, | |
| "loss": 0.989, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "eval_cer": 0.06105209914011128, | |
| "eval_loss": 0.23200440406799316, | |
| "eval_runtime": 15.3525, | |
| "eval_samples_per_second": 21.821, | |
| "eval_steps_per_second": 0.391, | |
| "eval_wer": 0.3251366120218579, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 26.91, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9518, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 26.91, | |
| "eval_cer": 0.05842185128983308, | |
| "eval_loss": 0.2413272261619568, | |
| "eval_runtime": 15.2385, | |
| "eval_samples_per_second": 21.984, | |
| "eval_steps_per_second": 0.394, | |
| "eval_wer": 0.3192818110850898, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 30.76, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9075, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 30.76, | |
| "eval_cer": 0.05933232169954476, | |
| "eval_loss": 0.23544833064079285, | |
| "eval_runtime": 15.1938, | |
| "eval_samples_per_second": 22.049, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.3200624512099922, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 34.61, | |
| "learning_rate": 7.059500000000001e-05, | |
| "loss": 0.878, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 34.61, | |
| "eval_cer": 0.057916034395548814, | |
| "eval_loss": 0.22777308523654938, | |
| "eval_runtime": 14.9728, | |
| "eval_samples_per_second": 22.374, | |
| "eval_steps_per_second": 0.401, | |
| "eval_wer": 0.3126463700234192, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 6.109500000000001e-05, | |
| "loss": 0.8563, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_cer": 0.054779969650986346, | |
| "eval_loss": 0.2326740324497223, | |
| "eval_runtime": 15.1749, | |
| "eval_samples_per_second": 22.076, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.2962529274004684, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 42.3, | |
| "learning_rate": 5.169000000000001e-05, | |
| "loss": 0.8084, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 42.3, | |
| "eval_cer": 0.05407182599898837, | |
| "eval_loss": 0.22712552547454834, | |
| "eval_runtime": 15.3083, | |
| "eval_samples_per_second": 21.884, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.2923497267759563, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 4.219000000000001e-05, | |
| "loss": 0.7845, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "eval_cer": 0.053667172483560954, | |
| "eval_loss": 0.23326420783996582, | |
| "eval_runtime": 15.1559, | |
| "eval_samples_per_second": 22.104, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.29508196721311475, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 49.99, | |
| "learning_rate": 3.269000000000001e-05, | |
| "loss": 0.7487, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 49.99, | |
| "eval_cer": 0.052453211937278706, | |
| "eval_loss": 0.22895006835460663, | |
| "eval_runtime": 15.373, | |
| "eval_samples_per_second": 21.791, | |
| "eval_steps_per_second": 0.39, | |
| "eval_wer": 0.2888368462138954, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 53.84, | |
| "learning_rate": 2.319e-05, | |
| "loss": 0.7182, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 53.84, | |
| "eval_cer": 0.05346484572584724, | |
| "eval_loss": 0.23406584560871124, | |
| "eval_runtime": 15.2056, | |
| "eval_samples_per_second": 22.031, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.28766588602654175, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 57.69, | |
| "learning_rate": 1.369e-05, | |
| "loss": 0.7095, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 57.69, | |
| "eval_cer": 0.05154274152756702, | |
| "eval_loss": 0.22908572852611542, | |
| "eval_runtime": 15.2684, | |
| "eval_samples_per_second": 21.941, | |
| "eval_steps_per_second": 0.393, | |
| "eval_wer": 0.2818110850897736, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 61.53, | |
| "learning_rate": 4.190000000000005e-06, | |
| "loss": 0.6953, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 61.53, | |
| "eval_cer": 0.05078401618614062, | |
| "eval_loss": 0.2248678207397461, | |
| "eval_runtime": 15.2139, | |
| "eval_samples_per_second": 22.019, | |
| "eval_steps_per_second": 0.394, | |
| "eval_wer": 0.2782982045277127, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 61.53, | |
| "step": 1600, | |
| "total_flos": 1.3126730002882698e+20, | |
| "train_loss": 1.2442097234725953, | |
| "train_runtime": 17605.3989, | |
| "train_samples_per_second": 11.633, | |
| "train_steps_per_second": 0.091 | |
| } | |
| ], | |
| "max_steps": 1600, | |
| "num_train_epochs": 62, | |
| "total_flos": 1.3126730002882698e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |