| { | |
| "best_metric": 0.452116459608078, | |
| "best_model_checkpoint": "./checkpoint-500", | |
| "epoch": 279.8695652173913, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 19.87, | |
| "learning_rate": 5.6240000000000004e-05, | |
| "loss": 6.1298, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 19.87, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.120361804962158, | |
| "eval_runtime": 16.2509, | |
| "eval_samples_per_second": 20.614, | |
| "eval_steps_per_second": 0.369, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 39.87, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7269, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 39.87, | |
| "eval_cer": 0.17546788062721294, | |
| "eval_loss": 0.6199544668197632, | |
| "eval_runtime": 17.4782, | |
| "eval_samples_per_second": 19.167, | |
| "eval_steps_per_second": 0.343, | |
| "eval_wer": 0.7591725214676034, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 59.87, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4643, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 59.87, | |
| "eval_cer": 0.12771876580677793, | |
| "eval_loss": 0.4795631468296051, | |
| "eval_runtime": 15.5884, | |
| "eval_samples_per_second": 21.49, | |
| "eval_steps_per_second": 0.385, | |
| "eval_wer": 0.5921155347384855, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 79.87, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1242, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 79.87, | |
| "eval_cer": 0.11451694486595852, | |
| "eval_loss": 0.463740736246109, | |
| "eval_runtime": 15.4298, | |
| "eval_samples_per_second": 21.711, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 0.5359094457455114, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 99.87, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 99.87, | |
| "eval_cer": 0.1099645928174001, | |
| "eval_loss": 0.452116459608078, | |
| "eval_runtime": 15.3829, | |
| "eval_samples_per_second": 21.777, | |
| "eval_steps_per_second": 0.39, | |
| "eval_wer": 0.5140515222482436, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 119.87, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8704, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 119.87, | |
| "eval_cer": 0.10450177035913, | |
| "eval_loss": 0.4736480712890625, | |
| "eval_runtime": 15.4613, | |
| "eval_samples_per_second": 21.667, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.49141295862607337, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 139.87, | |
| "learning_rate": 8e-05, | |
| "loss": 0.7908, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 139.87, | |
| "eval_cer": 0.11244309559939301, | |
| "eval_loss": 0.539383053779602, | |
| "eval_runtime": 15.4463, | |
| "eval_samples_per_second": 21.688, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.5249804839968775, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 159.87, | |
| "learning_rate": 6.936e-05, | |
| "loss": 0.7049, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 159.87, | |
| "eval_cer": 0.09848254931714719, | |
| "eval_loss": 0.48218029737472534, | |
| "eval_runtime": 15.323, | |
| "eval_samples_per_second": 21.863, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.47540983606557374, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 179.87, | |
| "learning_rate": 5.850285714285715e-05, | |
| "loss": 0.6299, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 179.87, | |
| "eval_cer": 0.1028325746079919, | |
| "eval_loss": 0.48903265595436096, | |
| "eval_runtime": 15.3742, | |
| "eval_samples_per_second": 21.79, | |
| "eval_steps_per_second": 0.39, | |
| "eval_wer": 0.4808743169398907, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 199.87, | |
| "learning_rate": 4.76457142857143e-05, | |
| "loss": 0.5832, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 199.87, | |
| "eval_cer": 0.10278199291856348, | |
| "eval_loss": 0.5233051180839539, | |
| "eval_runtime": 15.4506, | |
| "eval_samples_per_second": 21.682, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.4812646370023419, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 219.87, | |
| "learning_rate": 3.6788571428571434e-05, | |
| "loss": 0.5145, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 219.87, | |
| "eval_cer": 0.09939301972685888, | |
| "eval_loss": 0.5349759459495544, | |
| "eval_runtime": 15.4699, | |
| "eval_samples_per_second": 21.655, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.4781420765027322, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 239.87, | |
| "learning_rate": 2.5931428571428576e-05, | |
| "loss": 0.4604, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 239.87, | |
| "eval_cer": 0.09838138593829034, | |
| "eval_loss": 0.5222976803779602, | |
| "eval_runtime": 15.1676, | |
| "eval_samples_per_second": 22.087, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.4715066354410617, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 259.87, | |
| "learning_rate": 1.5074285714285721e-05, | |
| "loss": 0.4226, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 259.87, | |
| "eval_cer": 0.0952959028831563, | |
| "eval_loss": 0.5167204737663269, | |
| "eval_runtime": 15.5392, | |
| "eval_samples_per_second": 21.558, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.46252927400468385, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 279.87, | |
| "learning_rate": 4.217142857142858e-06, | |
| "loss": 0.3946, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 279.87, | |
| "eval_cer": 0.09504299443601416, | |
| "eval_loss": 0.5248004794120789, | |
| "eval_runtime": 15.5024, | |
| "eval_samples_per_second": 21.61, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 0.4613583138173302, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 279.87, | |
| "step": 1400, | |
| "total_flos": 1.252060110838857e+20, | |
| "train_loss": 1.2697014454432896, | |
| "train_runtime": 17182.2968, | |
| "train_samples_per_second": 10.429, | |
| "train_steps_per_second": 0.081 | |
| } | |
| ], | |
| "max_steps": 1400, | |
| "num_train_epochs": 280, | |
| "total_flos": 1.252060110838857e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |