| { | |
| "best_metric": 0.12257253378629684, | |
| "best_model_checkpoint": "./vit5_qqp/checkpoint-9855", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 9855, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15220700152207, | |
| "grad_norm": 0.5977054834365845, | |
| "learning_rate": 4.7463216641298836e-05, | |
| "loss": 0.2783, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.30441400304414, | |
| "grad_norm": 0.5528032779693604, | |
| "learning_rate": 4.492643328259767e-05, | |
| "loss": 0.1944, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.45662100456621, | |
| "grad_norm": 0.6017138361930847, | |
| "learning_rate": 4.2389649923896504e-05, | |
| "loss": 0.1803, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.60882800608828, | |
| "grad_norm": 0.4366663992404938, | |
| "learning_rate": 3.985286656519534e-05, | |
| "loss": 0.1719, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.76103500761035, | |
| "grad_norm": 0.4757032096385956, | |
| "learning_rate": 3.731608320649417e-05, | |
| "loss": 0.1655, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.91324200913242, | |
| "grad_norm": 0.45626300573349, | |
| "learning_rate": 3.4779299847793e-05, | |
| "loss": 0.1602, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.14291797578334808, | |
| "eval_runtime": 67.7907, | |
| "eval_samples_per_second": 193.065, | |
| "eval_steps_per_second": 5.369, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 1.06544901065449, | |
| "grad_norm": 0.44499117136001587, | |
| "learning_rate": 3.224251648909183e-05, | |
| "loss": 0.1472, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.21765601217656, | |
| "grad_norm": 0.3781445026397705, | |
| "learning_rate": 2.9705733130390666e-05, | |
| "loss": 0.1347, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 0.5086517930030823, | |
| "learning_rate": 2.71689497716895e-05, | |
| "loss": 0.1331, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.5220700152207, | |
| "grad_norm": 0.4261968731880188, | |
| "learning_rate": 2.463216641298833e-05, | |
| "loss": 0.1326, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.6742770167427703, | |
| "grad_norm": 0.4170425534248352, | |
| "learning_rate": 2.2095383054287164e-05, | |
| "loss": 0.131, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.82648401826484, | |
| "grad_norm": 0.5313445329666138, | |
| "learning_rate": 1.9558599695585997e-05, | |
| "loss": 0.129, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.9786910197869103, | |
| "grad_norm": 0.4916202425956726, | |
| "learning_rate": 1.702181633688483e-05, | |
| "loss": 0.1276, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.12813186645507812, | |
| "eval_runtime": 67.771, | |
| "eval_samples_per_second": 193.121, | |
| "eval_steps_per_second": 5.371, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 2.13089802130898, | |
| "grad_norm": 0.4530036449432373, | |
| "learning_rate": 1.4485032978183663e-05, | |
| "loss": 0.1129, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.2831050228310503, | |
| "grad_norm": 0.3842392861843109, | |
| "learning_rate": 1.1948249619482495e-05, | |
| "loss": 0.1092, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.43531202435312, | |
| "grad_norm": 0.4166797697544098, | |
| "learning_rate": 9.411466260781329e-06, | |
| "loss": 0.1086, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.5875190258751903, | |
| "grad_norm": 0.5072170495986938, | |
| "learning_rate": 6.874682902080162e-06, | |
| "loss": 0.1075, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 0.4664016664028168, | |
| "learning_rate": 4.337899543378996e-06, | |
| "loss": 0.1089, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.8919330289193304, | |
| "grad_norm": 0.48263782262802124, | |
| "learning_rate": 1.8011161846778284e-06, | |
| "loss": 0.1069, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.12257253378629684, | |
| "eval_runtime": 67.7838, | |
| "eval_samples_per_second": 193.084, | |
| "eval_steps_per_second": 5.37, | |
| "step": 9855 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 9855, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 2.11201621475328e+16, | |
| "train_batch_size": 36, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |