| { | |
| "best_metric": 0.837573385518591, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-25/checkpoint-768", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 768, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.8755850791931152, | |
| "learning_rate": 0.00020053170483737637, | |
| "loss": 0.4801, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7818003913894325, | |
| "eval_f1": 0.8092386655260906, | |
| "eval_loss": 0.42955365777015686, | |
| "eval_precision": 0.7188449848024316, | |
| "eval_recall": 0.9256360078277887, | |
| "eval_runtime": 31.124, | |
| "eval_samples_per_second": 32.836, | |
| "eval_steps_per_second": 1.028, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.4842429161071777, | |
| "learning_rate": 0.0001782504042998901, | |
| "loss": 0.4182, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8091976516634051, | |
| "eval_f1": 0.8232094288304623, | |
| "eval_loss": 0.3939443826675415, | |
| "eval_precision": 0.7668918918918919, | |
| "eval_recall": 0.8884540117416829, | |
| "eval_runtime": 32.508, | |
| "eval_samples_per_second": 31.438, | |
| "eval_steps_per_second": 0.984, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.5258917808532715, | |
| "learning_rate": 0.00015596910376240383, | |
| "loss": 0.3965, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.815068493150685, | |
| "eval_f1": 0.8325952170062001, | |
| "eval_loss": 0.39436766505241394, | |
| "eval_precision": 0.7605177993527508, | |
| "eval_recall": 0.9197651663405088, | |
| "eval_runtime": 29.4789, | |
| "eval_samples_per_second": 34.669, | |
| "eval_steps_per_second": 1.086, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.0768847465515137, | |
| "learning_rate": 0.00013368780322491758, | |
| "loss": 0.3841, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8140900195694716, | |
| "eval_f1": 0.8137254901960785, | |
| "eval_loss": 0.38306960463523865, | |
| "eval_precision": 0.8153241650294696, | |
| "eval_recall": 0.812133072407045, | |
| "eval_runtime": 32.8502, | |
| "eval_samples_per_second": 31.111, | |
| "eval_steps_per_second": 0.974, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.1314367055892944, | |
| "learning_rate": 0.00011140650268743131, | |
| "loss": 0.3779, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8326810176125244, | |
| "eval_f1": 0.8423963133640553, | |
| "eval_loss": 0.37976235151290894, | |
| "eval_precision": 0.7961672473867596, | |
| "eval_recall": 0.8943248532289628, | |
| "eval_runtime": 29.0986, | |
| "eval_samples_per_second": 35.122, | |
| "eval_steps_per_second": 1.1, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.076519012451172, | |
| "learning_rate": 8.912520214994506e-05, | |
| "loss": 0.3741, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.824853228962818, | |
| "eval_f1": 0.8397493285586393, | |
| "eval_loss": 0.3835048973560333, | |
| "eval_precision": 0.7739273927392739, | |
| "eval_recall": 0.9178082191780822, | |
| "eval_runtime": 28.8483, | |
| "eval_samples_per_second": 35.427, | |
| "eval_steps_per_second": 1.109, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.4347150325775146, | |
| "learning_rate": 6.684390161245879e-05, | |
| "loss": 0.3699, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8356164383561644, | |
| "eval_f1": 0.8469945355191257, | |
| "eval_loss": 0.3747766315937042, | |
| "eval_precision": 0.7921635434412265, | |
| "eval_recall": 0.9099804305283757, | |
| "eval_runtime": 28.9227, | |
| "eval_samples_per_second": 35.336, | |
| "eval_steps_per_second": 1.106, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.361301898956299, | |
| "learning_rate": 4.456260107497253e-05, | |
| "loss": 0.3677, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.837573385518591, | |
| "eval_f1": 0.8482632541133455, | |
| "eval_loss": 0.3733108639717102, | |
| "eval_precision": 0.7958833619210978, | |
| "eval_recall": 0.9080234833659491, | |
| "eval_runtime": 29.9904, | |
| "eval_samples_per_second": 34.078, | |
| "eval_steps_per_second": 1.067, | |
| "step": 768 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 960, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6872599698926333, | |
| "learning_rate": 0.00022281300537486262, | |
| "num_train_epochs": 10, | |
| "temperature": 15 | |
| } | |
| } | |