| { | |
| "best_metric": 0.5108951669698317, | |
| "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-3/checkpoint-8552", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 8552, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23386342376052385, | |
| "grad_norm": 9.916498184204102, | |
| "learning_rate": 1.0352460846585169e-05, | |
| "loss": 0.588, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4677268475210477, | |
| "grad_norm": 2.0442557334899902, | |
| "learning_rate": 9.709610570468354e-06, | |
| "loss": 0.5642, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7015902712815716, | |
| "grad_norm": 25.158737182617188, | |
| "learning_rate": 9.066760294351542e-06, | |
| "loss": 0.5358, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9354536950420954, | |
| "grad_norm": 21.2142276763916, | |
| "learning_rate": 8.423910018234727e-06, | |
| "loss": 0.5288, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6297594904899597, | |
| "eval_matthews_correlation": 0.4181278600904661, | |
| "eval_runtime": 0.8372, | |
| "eval_samples_per_second": 1245.804, | |
| "eval_steps_per_second": 78.833, | |
| "step": 2138 | |
| }, | |
| { | |
| "epoch": 1.1693171188026192, | |
| "grad_norm": 40.90157699584961, | |
| "learning_rate": 7.781059742117914e-06, | |
| "loss": 0.4866, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4031805425631432, | |
| "grad_norm": 0.11679931730031967, | |
| "learning_rate": 7.138209466001099e-06, | |
| "loss": 0.4601, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.637043966323667, | |
| "grad_norm": 9.670472145080566, | |
| "learning_rate": 6.495359189884285e-06, | |
| "loss": 0.4758, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.8709073900841908, | |
| "grad_norm": 1.609247088432312, | |
| "learning_rate": 5.852508913767472e-06, | |
| "loss": 0.4517, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.9118176102638245, | |
| "eval_matthews_correlation": 0.4622674649892012, | |
| "eval_runtime": 0.7476, | |
| "eval_samples_per_second": 1395.206, | |
| "eval_steps_per_second": 88.287, | |
| "step": 4276 | |
| }, | |
| { | |
| "epoch": 2.1047708138447145, | |
| "grad_norm": 0.1278693974018097, | |
| "learning_rate": 5.209658637650659e-06, | |
| "loss": 0.3716, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.3386342376052385, | |
| "grad_norm": 0.046489011496305466, | |
| "learning_rate": 4.566808361533844e-06, | |
| "loss": 0.3199, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.5724976613657624, | |
| "grad_norm": 8.200950622558594, | |
| "learning_rate": 3.923958085417031e-06, | |
| "loss": 0.3667, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.8063610851262863, | |
| "grad_norm": 0.06558237969875336, | |
| "learning_rate": 3.281107809300217e-06, | |
| "loss": 0.3323, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.9285444021224976, | |
| "eval_matthews_correlation": 0.5015847043424765, | |
| "eval_runtime": 0.7214, | |
| "eval_samples_per_second": 1445.788, | |
| "eval_steps_per_second": 91.488, | |
| "step": 6414 | |
| }, | |
| { | |
| "epoch": 3.0402245088868103, | |
| "grad_norm": 0.10314546525478363, | |
| "learning_rate": 2.6382575331834036e-06, | |
| "loss": 0.325, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.2740879326473338, | |
| "grad_norm": 0.09269363433122635, | |
| "learning_rate": 1.99540725706659e-06, | |
| "loss": 0.2483, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.5079513564078577, | |
| "grad_norm": 0.033062130212783813, | |
| "learning_rate": 1.3525569809497761e-06, | |
| "loss": 0.2065, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.7418147801683816, | |
| "grad_norm": 2.666268825531006, | |
| "learning_rate": 7.097067048329624e-07, | |
| "loss": 0.2449, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.9756782039289056, | |
| "grad_norm": 0.12107378989458084, | |
| "learning_rate": 6.685642871614863e-08, | |
| "loss": 0.2523, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.092543125152588, | |
| "eval_matthews_correlation": 0.5108951669698317, | |
| "eval_runtime": 0.8843, | |
| "eval_samples_per_second": 1179.471, | |
| "eval_steps_per_second": 74.636, | |
| "step": 8552 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8552, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 138564521149608.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.0995311122701982e-05, | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 4, | |
| "seed": 21 | |
| } | |
| } | |