| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.416856492027335, | |
| "eval_steps": 200, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11389521640091116, | |
| "eval_loss": 3.107494354248047, | |
| "eval_runtime": 174.2934, | |
| "eval_samples_per_second": 40.288, | |
| "eval_steps_per_second": 5.037, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22779043280182232, | |
| "eval_loss": 1.6248737573623657, | |
| "eval_runtime": 174.1392, | |
| "eval_samples_per_second": 40.324, | |
| "eval_steps_per_second": 5.042, | |
| "eval_wer": 0.8958131253404908, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2847380410022779, | |
| "grad_norm": 3.3784055709838867, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.6043, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3416856492027335, | |
| "eval_loss": 1.1246436834335327, | |
| "eval_runtime": 174.2087, | |
| "eval_samples_per_second": 40.308, | |
| "eval_steps_per_second": 5.04, | |
| "eval_wer": 0.8675772997249498, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "eval_loss": 0.9142205715179443, | |
| "eval_runtime": 174.1428, | |
| "eval_samples_per_second": 40.323, | |
| "eval_steps_per_second": 5.042, | |
| "eval_wer": 0.7093902390411778, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 2.1683263778686523, | |
| "learning_rate": 0.0002730545454545454, | |
| "loss": 1.0665, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "eval_loss": 0.7569850087165833, | |
| "eval_runtime": 173.8805, | |
| "eval_samples_per_second": 40.384, | |
| "eval_steps_per_second": 5.049, | |
| "eval_wer": 0.5934306860309067, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "eval_loss": 0.6979650259017944, | |
| "eval_runtime": 174.537, | |
| "eval_samples_per_second": 40.232, | |
| "eval_steps_per_second": 5.03, | |
| "eval_wer": 0.5284816433914881, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7972665148063781, | |
| "eval_loss": 0.6941590905189514, | |
| "eval_runtime": 174.9006, | |
| "eval_samples_per_second": 40.149, | |
| "eval_steps_per_second": 5.02, | |
| "eval_wer": 0.5388990021127041, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8542141230068337, | |
| "grad_norm": 2.244067907333374, | |
| "learning_rate": 0.0002459454545454545, | |
| "loss": 0.8811, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "eval_loss": 0.6152143478393555, | |
| "eval_runtime": 175.8479, | |
| "eval_samples_per_second": 39.932, | |
| "eval_steps_per_second": 4.993, | |
| "eval_wer": 0.4861611235865478, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.0250569476082005, | |
| "eval_loss": 0.5941163301467896, | |
| "eval_runtime": 174.9993, | |
| "eval_samples_per_second": 40.126, | |
| "eval_steps_per_second": 5.017, | |
| "eval_wer": 0.4534607156619142, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 2.146209478378296, | |
| "learning_rate": 0.00021878181818181817, | |
| "loss": 0.7654, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "eval_loss": 0.5642105340957642, | |
| "eval_runtime": 174.8742, | |
| "eval_samples_per_second": 40.155, | |
| "eval_steps_per_second": 5.021, | |
| "eval_wer": 0.43286517227175486, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.2528473804100229, | |
| "eval_loss": 0.5727251768112183, | |
| "eval_runtime": 175.3644, | |
| "eval_samples_per_second": 40.042, | |
| "eval_steps_per_second": 5.007, | |
| "eval_wer": 0.43133711582667855, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "eval_loss": 0.5466997027397156, | |
| "eval_runtime": 175.0585, | |
| "eval_samples_per_second": 40.112, | |
| "eval_steps_per_second": 5.015, | |
| "eval_wer": 0.4316825894577393, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.4236902050113895, | |
| "grad_norm": 1.5887110233306885, | |
| "learning_rate": 0.00019167272727272725, | |
| "loss": 0.6896, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4806378132118452, | |
| "eval_loss": 0.5658189654350281, | |
| "eval_runtime": 178.0734, | |
| "eval_samples_per_second": 39.433, | |
| "eval_steps_per_second": 4.931, | |
| "eval_wer": 0.43980121978766656, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "eval_loss": 0.6008076071739197, | |
| "eval_runtime": 175.5831, | |
| "eval_samples_per_second": 39.992, | |
| "eval_steps_per_second": 5.0, | |
| "eval_wer": 0.4445182635963805, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 3.4384372234344482, | |
| "learning_rate": 0.0001645090909090909, | |
| "loss": 0.7558, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "eval_loss": 0.7089642286300659, | |
| "eval_runtime": 175.5809, | |
| "eval_samples_per_second": 39.993, | |
| "eval_steps_per_second": 5.001, | |
| "eval_wer": 0.5020927729573872, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "eval_loss": 0.8049871921539307, | |
| "eval_runtime": 175.5153, | |
| "eval_samples_per_second": 40.008, | |
| "eval_steps_per_second": 5.002, | |
| "eval_wer": 0.5291327283115641, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.9362186788154898, | |
| "eval_loss": 0.824817955493927, | |
| "eval_runtime": 175.9347, | |
| "eval_samples_per_second": 39.913, | |
| "eval_steps_per_second": 4.99, | |
| "eval_wer": 0.5275382346297453, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.9931662870159452, | |
| "grad_norm": 3.6540627479553223, | |
| "learning_rate": 0.00013723636363636361, | |
| "loss": 0.9823, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "eval_loss": 0.7944601774215698, | |
| "eval_runtime": 176.7964, | |
| "eval_samples_per_second": 39.718, | |
| "eval_steps_per_second": 4.966, | |
| "eval_wer": 0.5072748774232982, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.164009111617312, | |
| "eval_loss": 0.7690120935440063, | |
| "eval_runtime": 177.7491, | |
| "eval_samples_per_second": 39.505, | |
| "eval_steps_per_second": 4.94, | |
| "eval_wer": 0.4982128383316281, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 12.11134147644043, | |
| "learning_rate": 0.00011001818181818181, | |
| "loss": 0.9181, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.8668798804283142, | |
| "eval_runtime": 176.748, | |
| "eval_samples_per_second": 39.729, | |
| "eval_steps_per_second": 4.968, | |
| "eval_wer": 0.6929802415657927, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.3917995444191344, | |
| "eval_loss": 0.88874351978302, | |
| "eval_runtime": 177.153, | |
| "eval_samples_per_second": 39.638, | |
| "eval_steps_per_second": 4.956, | |
| "eval_wer": 0.6311271741585724, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.5056947608200457, | |
| "eval_loss": 1.150854468345642, | |
| "eval_runtime": 176.8956, | |
| "eval_samples_per_second": 39.696, | |
| "eval_steps_per_second": 4.963, | |
| "eval_wer": 0.8436200321556225, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.562642369020501, | |
| "grad_norm": 6.307694911956787, | |
| "learning_rate": 8.285454545454545e-05, | |
| "loss": 1.0805, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.619589977220957, | |
| "eval_loss": 1.2166552543640137, | |
| "eval_runtime": 176.6784, | |
| "eval_samples_per_second": 39.745, | |
| "eval_steps_per_second": 4.969, | |
| "eval_wer": 0.9458935143969492, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "eval_loss": 1.2323366403579712, | |
| "eval_runtime": 177.2283, | |
| "eval_samples_per_second": 39.621, | |
| "eval_steps_per_second": 4.954, | |
| "eval_wer": 0.9307059620776252, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "grad_norm": 0.8214463591575623, | |
| "learning_rate": 5.569090909090908e-05, | |
| "loss": 1.3977, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "eval_loss": 1.65969979763031, | |
| "eval_runtime": 176.9601, | |
| "eval_samples_per_second": 39.681, | |
| "eval_steps_per_second": 4.962, | |
| "eval_wer": 0.9527896995708155, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.9612756264236904, | |
| "eval_loss": 2.022368907928467, | |
| "eval_runtime": 176.5944, | |
| "eval_samples_per_second": 39.763, | |
| "eval_steps_per_second": 4.972, | |
| "eval_wer": 0.9162359319151198, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.075170842824601, | |
| "eval_loss": 2.105426788330078, | |
| "eval_runtime": 176.3683, | |
| "eval_samples_per_second": 39.814, | |
| "eval_steps_per_second": 4.978, | |
| "eval_wer": 0.9103761676344357, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 3.132118451025057, | |
| "grad_norm": 0.6184139251708984, | |
| "learning_rate": 2.8472727272727272e-05, | |
| "loss": 2.1028, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.1890660592255125, | |
| "eval_loss": 2.248642683029175, | |
| "eval_runtime": 175.7372, | |
| "eval_samples_per_second": 39.957, | |
| "eval_steps_per_second": 4.996, | |
| "eval_wer": 0.9280750475026243, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.3029612756264237, | |
| "eval_loss": 2.2446603775024414, | |
| "eval_runtime": 176.6127, | |
| "eval_samples_per_second": 39.759, | |
| "eval_steps_per_second": 4.971, | |
| "eval_wer": 0.9215774857492127, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "grad_norm": 0.6574464440345764, | |
| "learning_rate": 1.3636363636363634e-06, | |
| "loss": 2.2911, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "eval_loss": 2.2309672832489014, | |
| "eval_runtime": 176.1017, | |
| "eval_samples_per_second": 39.875, | |
| "eval_steps_per_second": 4.986, | |
| "eval_wer": 0.9239559388245924, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "step": 6000, | |
| "total_flos": 1.116875973325552e+19, | |
| "train_loss": 1.37794411722819, | |
| "train_runtime": 8509.4041, | |
| "train_samples_per_second": 11.282, | |
| "train_steps_per_second": 0.705 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 1.116875973325552e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |