| { | |
| "best_metric": 0.5440313111545988, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-7/checkpoint-292", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 876, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4084218442440033, | |
| "learning_rate": 0.0002846359367487843, | |
| "loss": 0.0564, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5420743639921722, | |
| "eval_f1": 0.21212121212121213, | |
| "eval_loss": 0.042483557015657425, | |
| "eval_precision": 0.7590361445783133, | |
| "eval_recall": 0.1232876712328767, | |
| "eval_runtime": 28.2636, | |
| "eval_samples_per_second": 36.16, | |
| "eval_steps_per_second": 1.132, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.22884152829647064, | |
| "learning_rate": 0.00022770874939902745, | |
| "loss": 0.0455, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5440313111545988, | |
| "eval_f1": 0.21283783783783783, | |
| "eval_loss": 0.04199887067079544, | |
| "eval_precision": 0.7777777777777778, | |
| "eval_recall": 0.1232876712328767, | |
| "eval_runtime": 27.8033, | |
| "eval_samples_per_second": 36.758, | |
| "eval_steps_per_second": 1.151, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.41705241799354553, | |
| "learning_rate": 0.0001707815620492706, | |
| "loss": 0.0445, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5322896281800391, | |
| "eval_f1": 0.1758620689655172, | |
| "eval_loss": 0.0437779501080513, | |
| "eval_precision": 0.7391304347826086, | |
| "eval_recall": 0.09980430528375733, | |
| "eval_runtime": 28.2813, | |
| "eval_samples_per_second": 36.137, | |
| "eval_steps_per_second": 1.131, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.17889831960201263, | |
| "learning_rate": 0.00011385437469951372, | |
| "loss": 0.0435, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5332681017612525, | |
| "eval_f1": 0.17900172117039587, | |
| "eval_loss": 0.04091888666152954, | |
| "eval_precision": 0.7428571428571429, | |
| "eval_recall": 0.10176125244618395, | |
| "eval_runtime": 28.0708, | |
| "eval_samples_per_second": 36.408, | |
| "eval_steps_per_second": 1.14, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.35064178705215454, | |
| "learning_rate": 5.692718734975686e-05, | |
| "loss": 0.0431, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5322896281800391, | |
| "eval_f1": 0.1758620689655172, | |
| "eval_loss": 0.04131113365292549, | |
| "eval_precision": 0.7391304347826086, | |
| "eval_recall": 0.09980430528375733, | |
| "eval_runtime": 27.8905, | |
| "eval_samples_per_second": 36.643, | |
| "eval_steps_per_second": 1.147, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.22150251269340515, | |
| "learning_rate": 0.0, | |
| "loss": 0.0427, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5352250489236791, | |
| "eval_f1": 0.18524871355060032, | |
| "eval_loss": 0.04079839587211609, | |
| "eval_precision": 0.75, | |
| "eval_recall": 0.10567514677103718, | |
| "eval_runtime": 27.877, | |
| "eval_samples_per_second": 36.661, | |
| "eval_steps_per_second": 1.148, | |
| "step": 876 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 876, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 1414171183680.0, | |
| "train_batch_size": 21, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.04464648890006717, | |
| "learning_rate": 0.0003415631240985412, | |
| "num_train_epochs": 6, | |
| "per_device_train_batch_size": 21, | |
| "temperature": 2 | |
| } | |
| } | |