| { | |
| "best_metric": Infinity, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.8174178762414055, | |
| "eval_steps": 100, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15278838808250572, | |
| "grad_norm": 10.146588325500488, | |
| "learning_rate": 0.000285, | |
| "loss": 14.705, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15278838808250572, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.1562, | |
| "eval_samples_per_second": 14.059, | |
| "eval_steps_per_second": 3.515, | |
| "eval_wer": 1.0003503035964503, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30557677616501144, | |
| "grad_norm": 5.455625057220459, | |
| "learning_rate": 0.00028812499999999997, | |
| "loss": 6.0879, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30557677616501144, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.1167, | |
| "eval_samples_per_second": 14.067, | |
| "eval_steps_per_second": 3.517, | |
| "eval_wer": 1.0203176085941148, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45836516424751717, | |
| "grad_norm": 5.808646202087402, | |
| "learning_rate": 0.00027562499999999994, | |
| "loss": 5.4042, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.45836516424751717, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.2917, | |
| "eval_samples_per_second": 14.029, | |
| "eval_steps_per_second": 3.507, | |
| "eval_wer": 1.0159971975712283, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6111535523300229, | |
| "grad_norm": 2.3630435466766357, | |
| "learning_rate": 0.00026312499999999996, | |
| "loss": 2.8156, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6111535523300229, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.5406, | |
| "eval_samples_per_second": 13.976, | |
| "eval_steps_per_second": 3.494, | |
| "eval_wer": 0.6057916861279776, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7639419404125286, | |
| "grad_norm": 11.420530319213867, | |
| "learning_rate": 0.000250625, | |
| "loss": 1.9111, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7639419404125286, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.6804, | |
| "eval_samples_per_second": 13.946, | |
| "eval_steps_per_second": 3.487, | |
| "eval_wer": 0.5723960765997198, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9167303284950343, | |
| "grad_norm": 6.228967189788818, | |
| "learning_rate": 0.00023812499999999997, | |
| "loss": 1.82, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9167303284950343, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.0545, | |
| "eval_samples_per_second": 14.081, | |
| "eval_steps_per_second": 3.52, | |
| "eval_wer": 0.5643390938813638, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0687547746371275, | |
| "grad_norm": 5.563028812408447, | |
| "learning_rate": 0.00022562499999999997, | |
| "loss": 1.6294, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0687547746371275, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.1677, | |
| "eval_samples_per_second": 14.056, | |
| "eval_steps_per_second": 3.514, | |
| "eval_wer": 0.5685427370387669, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2215431627196334, | |
| "grad_norm": 2.7232375144958496, | |
| "learning_rate": 0.000213125, | |
| "loss": 1.6856, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2215431627196334, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.032, | |
| "eval_samples_per_second": 13.872, | |
| "eval_steps_per_second": 3.468, | |
| "eval_wer": 0.5530126109294722, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.374331550802139, | |
| "grad_norm": 2.6046690940856934, | |
| "learning_rate": 0.00020062499999999996, | |
| "loss": 1.6363, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.374331550802139, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.9815, | |
| "eval_samples_per_second": 13.883, | |
| "eval_steps_per_second": 3.471, | |
| "eval_wer": 0.5500934142923868, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.5271199388846448, | |
| "grad_norm": 2.4681789875030518, | |
| "learning_rate": 0.00018812499999999998, | |
| "loss": 1.5114, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5271199388846448, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.3495, | |
| "eval_samples_per_second": 14.017, | |
| "eval_steps_per_second": 3.504, | |
| "eval_wer": 0.5416861279775805, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6799083269671504, | |
| "grad_norm": 2.219026803970337, | |
| "learning_rate": 0.000175625, | |
| "loss": 1.5417, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6799083269671504, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.2544, | |
| "eval_samples_per_second": 14.037, | |
| "eval_steps_per_second": 3.509, | |
| "eval_wer": 0.5358477347034096, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8326967150496563, | |
| "grad_norm": 7.856187343597412, | |
| "learning_rate": 0.00016312499999999997, | |
| "loss": 1.6518, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8326967150496563, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.3581, | |
| "eval_samples_per_second": 13.804, | |
| "eval_steps_per_second": 3.451, | |
| "eval_wer": 0.5337459131247081, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.985485103132162, | |
| "grad_norm": 3.3855655193328857, | |
| "learning_rate": 0.000150625, | |
| "loss": 1.4795, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.985485103132162, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.8957, | |
| "eval_samples_per_second": 13.901, | |
| "eval_steps_per_second": 3.475, | |
| "eval_wer": 0.5291919663708547, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.137509549274255, | |
| "grad_norm": 3.592418670654297, | |
| "learning_rate": 0.00013812499999999998, | |
| "loss": 1.5822, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.137509549274255, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.4473, | |
| "eval_samples_per_second": 13.996, | |
| "eval_steps_per_second": 3.499, | |
| "eval_wer": 0.5277907519850538, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.290297937356761, | |
| "grad_norm": 5.870262622833252, | |
| "learning_rate": 0.000125625, | |
| "loss": 1.4938, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.290297937356761, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.3633, | |
| "eval_samples_per_second": 14.014, | |
| "eval_steps_per_second": 3.503, | |
| "eval_wer": 0.5193834656702475, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4430863254392667, | |
| "grad_norm": 2.455545663833618, | |
| "learning_rate": 0.00011312499999999999, | |
| "loss": 1.5701, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4430863254392667, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.1226, | |
| "eval_samples_per_second": 13.853, | |
| "eval_steps_per_second": 3.463, | |
| "eval_wer": 0.5352638953759925, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.5958747135217726, | |
| "grad_norm": 2.66916823387146, | |
| "learning_rate": 0.00010062499999999998, | |
| "loss": 1.47, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.5958747135217726, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.22, | |
| "eval_samples_per_second": 13.833, | |
| "eval_steps_per_second": 3.458, | |
| "eval_wer": 0.5183325548808968, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.748663101604278, | |
| "grad_norm": 7.257411956787109, | |
| "learning_rate": 8.8125e-05, | |
| "loss": 1.4109, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.748663101604278, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.3542, | |
| "eval_samples_per_second": 14.016, | |
| "eval_steps_per_second": 3.504, | |
| "eval_wer": 0.5342129845866418, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.901451489686784, | |
| "grad_norm": 3.3290886878967285, | |
| "learning_rate": 7.5625e-05, | |
| "loss": 1.3993, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.901451489686784, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.5447, | |
| "eval_samples_per_second": 13.975, | |
| "eval_steps_per_second": 3.494, | |
| "eval_wer": 0.5175151798225128, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.053475935828877, | |
| "grad_norm": 2.5621745586395264, | |
| "learning_rate": 6.312499999999999e-05, | |
| "loss": 1.4848, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.053475935828877, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.0221, | |
| "eval_samples_per_second": 13.874, | |
| "eval_steps_per_second": 3.469, | |
| "eval_wer": 0.5177487155534797, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2062643239113826, | |
| "grad_norm": 5.282195568084717, | |
| "learning_rate": 5.0625e-05, | |
| "loss": 1.4331, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.2062643239113826, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.1067, | |
| "eval_samples_per_second": 13.856, | |
| "eval_steps_per_second": 3.464, | |
| "eval_wer": 0.5286081270434376, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.3590527119938884, | |
| "grad_norm": 4.0164008140563965, | |
| "learning_rate": 3.812499999999999e-05, | |
| "loss": 1.4392, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3590527119938884, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.662, | |
| "eval_samples_per_second": 13.95, | |
| "eval_steps_per_second": 3.488, | |
| "eval_wer": 0.5212517515179822, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.5118411000763943, | |
| "grad_norm": 8.766704559326172, | |
| "learning_rate": 2.5625e-05, | |
| "loss": 1.3324, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5118411000763943, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 65.6317, | |
| "eval_samples_per_second": 13.957, | |
| "eval_steps_per_second": 3.489, | |
| "eval_wer": 0.5161139654367118, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.6646294881588997, | |
| "grad_norm": 2.9708027839660645, | |
| "learning_rate": 1.3124999999999999e-05, | |
| "loss": 1.501, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.6646294881588997, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.0549, | |
| "eval_samples_per_second": 13.867, | |
| "eval_steps_per_second": 3.467, | |
| "eval_wer": 0.5159971975712284, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.8174178762414055, | |
| "grad_norm": 3.6458466053009033, | |
| "learning_rate": 6.249999999999999e-07, | |
| "loss": 1.3526, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8174178762414055, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 66.6457, | |
| "eval_samples_per_second": 13.744, | |
| "eval_steps_per_second": 3.436, | |
| "eval_wer": 0.5163475011676787, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8174178762414055, | |
| "step": 2500, | |
| "total_flos": 1.2788709685256737e+19, | |
| "train_loss": 2.4539583984375, | |
| "train_runtime": 5128.7055, | |
| "train_samples_per_second": 3.9, | |
| "train_steps_per_second": 0.487 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2788709685256737e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |