| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 39.926140859931415, | |
| "eval_steps": 1500, | |
| "global_step": 18920, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 0.0002917131696428571, | |
| "loss": 6.3673, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_cer": 0.16064462437366714, | |
| "eval_loss": 0.6103929877281189, | |
| "eval_runtime": 143.9739, | |
| "eval_samples_per_second": 26.185, | |
| "eval_steps_per_second": 3.278, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "learning_rate": 0.00026660156249999997, | |
| "loss": 0.656, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "eval_cer": 0.112913292993741, | |
| "eval_loss": 0.4318251311779022, | |
| "eval_runtime": 143.6016, | |
| "eval_samples_per_second": 26.253, | |
| "eval_steps_per_second": 3.287, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 0.00024148995535714285, | |
| "loss": 0.4729, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "eval_cer": 0.10276193283284499, | |
| "eval_loss": 0.4009631276130676, | |
| "eval_runtime": 144.1302, | |
| "eval_samples_per_second": 26.157, | |
| "eval_steps_per_second": 3.275, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "learning_rate": 0.0002163783482142857, | |
| "loss": 0.3789, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "eval_cer": 0.09767324930214817, | |
| "eval_loss": 0.3866880536079407, | |
| "eval_runtime": 145.6008, | |
| "eval_samples_per_second": 25.893, | |
| "eval_steps_per_second": 3.242, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "learning_rate": 0.00019128348214285712, | |
| "loss": 0.3166, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "eval_cer": 0.0936248417913553, | |
| "eval_loss": 0.38572949171066284, | |
| "eval_runtime": 144.2607, | |
| "eval_samples_per_second": 26.133, | |
| "eval_steps_per_second": 3.272, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "learning_rate": 0.00016618861607142855, | |
| "loss": 0.267, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "eval_cer": 0.0911715241777485, | |
| "eval_loss": 0.389119952917099, | |
| "eval_runtime": 143.5612, | |
| "eval_samples_per_second": 26.261, | |
| "eval_steps_per_second": 3.288, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 22.16, | |
| "learning_rate": 0.00014107700892857143, | |
| "loss": 0.2286, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 22.16, | |
| "eval_cer": 0.09095480000693518, | |
| "eval_loss": 0.40744006633758545, | |
| "eval_runtime": 143.443, | |
| "eval_samples_per_second": 26.282, | |
| "eval_steps_per_second": 3.291, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "learning_rate": 0.00011598214285714285, | |
| "loss": 0.1967, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "eval_cer": 0.08783397194722332, | |
| "eval_loss": 0.4078831076622009, | |
| "eval_runtime": 142.9582, | |
| "eval_samples_per_second": 26.371, | |
| "eval_steps_per_second": 3.302, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 28.49, | |
| "learning_rate": 9.08705357142857e-05, | |
| "loss": 0.1712, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 28.49, | |
| "eval_cer": 0.08651628898867833, | |
| "eval_loss": 0.42893365025520325, | |
| "eval_runtime": 143.8273, | |
| "eval_samples_per_second": 26.212, | |
| "eval_steps_per_second": 3.282, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "learning_rate": 6.577566964285713e-05, | |
| "loss": 0.1493, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "eval_cer": 0.08501655772665014, | |
| "eval_loss": 0.4456014931201935, | |
| "eval_runtime": 144.3831, | |
| "eval_samples_per_second": 26.111, | |
| "eval_steps_per_second": 3.269, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 34.82, | |
| "learning_rate": 4.068080357142857e-05, | |
| "loss": 0.1333, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 34.82, | |
| "eval_cer": 0.08429703347954991, | |
| "eval_loss": 0.457296758890152, | |
| "eval_runtime": 144.9279, | |
| "eval_samples_per_second": 26.013, | |
| "eval_steps_per_second": 3.257, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 37.98, | |
| "learning_rate": 1.556919642857143e-05, | |
| "loss": 0.1191, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 37.98, | |
| "eval_cer": 0.08327409539331103, | |
| "eval_loss": 0.4633271098136902, | |
| "eval_runtime": 146.5025, | |
| "eval_samples_per_second": 25.733, | |
| "eval_steps_per_second": 3.222, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 39.93, | |
| "step": 18920, | |
| "total_flos": 1.5462553297546432e+20, | |
| "train_loss": 0.7552054080600215, | |
| "train_runtime": 32439.3262, | |
| "train_samples_per_second": 37.392, | |
| "train_steps_per_second": 0.583 | |
| } | |
| ], | |
| "logging_steps": 1500, | |
| "max_steps": 18920, | |
| "num_train_epochs": 40, | |
| "save_steps": 1500, | |
| "total_flos": 1.5462553297546432e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |