| { | |
| "best_metric": 0.6961374878883362, | |
| "best_model_checkpoint": "distilbert_lda_5_v1_rte/checkpoint-30", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.3214802742004395, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.7504, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.7092382311820984, | |
| "eval_runtime": 0.1294, | |
| "eval_samples_per_second": 2141.351, | |
| "eval_steps_per_second": 15.461, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0456618070602417, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.6796, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.516245487364621, | |
| "eval_loss": 0.7038653492927551, | |
| "eval_runtime": 0.1272, | |
| "eval_samples_per_second": 2177.108, | |
| "eval_steps_per_second": 15.719, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.003218650817871, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.6429, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5090252707581228, | |
| "eval_loss": 0.6961374878883362, | |
| "eval_runtime": 0.129, | |
| "eval_samples_per_second": 2146.522, | |
| "eval_steps_per_second": 15.498, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.37013840675354, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5697, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5415162454873647, | |
| "eval_loss": 0.7390656471252441, | |
| "eval_runtime": 0.1247, | |
| "eval_samples_per_second": 2221.135, | |
| "eval_steps_per_second": 16.037, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.7822569608688354, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4539, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5306859205776173, | |
| "eval_loss": 0.8573698997497559, | |
| "eval_runtime": 0.1268, | |
| "eval_samples_per_second": 2185.207, | |
| "eval_steps_per_second": 15.778, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.516984939575195, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3105, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5523465703971119, | |
| "eval_loss": 1.0425331592559814, | |
| "eval_runtime": 0.126, | |
| "eval_samples_per_second": 2197.919, | |
| "eval_steps_per_second": 15.869, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 5.505280017852783, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.2107, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5054151624548736, | |
| "eval_loss": 1.4041368961334229, | |
| "eval_runtime": 0.1276, | |
| "eval_samples_per_second": 2171.224, | |
| "eval_steps_per_second": 15.677, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.022568941116333, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.1354, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.51985559566787, | |
| "eval_loss": 1.4544810056686401, | |
| "eval_runtime": 0.1345, | |
| "eval_samples_per_second": 2058.888, | |
| "eval_steps_per_second": 14.866, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 80, | |
| "total_flos": 1319375290613760.0, | |
| "train_loss": 0.46914870738983155, | |
| "train_runtime": 32.4978, | |
| "train_samples_per_second": 3831.025, | |
| "train_steps_per_second": 15.386 | |
| } | |
| ], | |
| "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": 1319375290613760.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |