| { | |
| "best_global_step": 50, | |
| "best_metric": 0.6918716430664062, | |
| "best_model_checkpoint": "bert_base_train_book_ent_15p_s_init_rte/checkpoint-50", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0603665113449097, | |
| "learning_rate": 4.91e-05, | |
| "loss": 0.7023, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6951573491096497, | |
| "eval_runtime": 0.2405, | |
| "eval_samples_per_second": 1151.977, | |
| "eval_steps_per_second": 8.318, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.376626968383789, | |
| "learning_rate": 4.8100000000000004e-05, | |
| "loss": 0.7018, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6921254396438599, | |
| "eval_runtime": 0.2574, | |
| "eval_samples_per_second": 1076.002, | |
| "eval_steps_per_second": 7.769, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.1199655532836914, | |
| "learning_rate": 4.71e-05, | |
| "loss": 0.6966, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.694311261177063, | |
| "eval_runtime": 0.2546, | |
| "eval_samples_per_second": 1088.134, | |
| "eval_steps_per_second": 7.857, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.0557060241699219, | |
| "learning_rate": 4.61e-05, | |
| "loss": 0.6977, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6960176825523376, | |
| "eval_runtime": 0.24, | |
| "eval_samples_per_second": 1154.243, | |
| "eval_steps_per_second": 8.334, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.9181549549102783, | |
| "learning_rate": 4.5100000000000005e-05, | |
| "loss": 0.6969, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6918716430664062, | |
| "eval_runtime": 0.2419, | |
| "eval_samples_per_second": 1145.205, | |
| "eval_steps_per_second": 8.269, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6541225910186768, | |
| "learning_rate": 4.41e-05, | |
| "loss": 0.7016, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6975829005241394, | |
| "eval_runtime": 0.24, | |
| "eval_samples_per_second": 1154.058, | |
| "eval_steps_per_second": 8.333, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.42498037219047546, | |
| "learning_rate": 4.3100000000000004e-05, | |
| "loss": 0.6981, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6936765909194946, | |
| "eval_runtime": 0.2461, | |
| "eval_samples_per_second": 1125.367, | |
| "eval_steps_per_second": 8.125, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.4852175712585449, | |
| "learning_rate": 4.21e-05, | |
| "loss": 0.6952, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6927741169929504, | |
| "eval_runtime": 0.2448, | |
| "eval_samples_per_second": 1131.61, | |
| "eval_steps_per_second": 8.17, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.571114718914032, | |
| "learning_rate": 4.11e-05, | |
| "loss": 0.6986, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6929998397827148, | |
| "eval_runtime": 0.2444, | |
| "eval_samples_per_second": 1133.421, | |
| "eval_steps_per_second": 8.184, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5091403722763062, | |
| "learning_rate": 4.0100000000000006e-05, | |
| "loss": 0.6978, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6923933625221252, | |
| "eval_runtime": 0.2389, | |
| "eval_samples_per_second": 1159.44, | |
| "eval_steps_per_second": 8.371, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 100, | |
| "total_flos": 3275732639232000.0, | |
| "train_loss": 0.698671703338623, | |
| "train_runtime": 72.3718, | |
| "train_samples_per_second": 1720.284, | |
| "train_steps_per_second": 6.909 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3275732639232000.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |