| { | |
| "best_metric": 0.8356164383561644, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-47/checkpoint-384", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 384, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 10.090946197509766, | |
| "learning_rate": 0.0002726931435690813, | |
| "loss": 0.5743, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7964774951076321, | |
| "eval_f1": 0.8188153310104529, | |
| "eval_loss": 0.4561440944671631, | |
| "eval_precision": 0.7378335949764521, | |
| "eval_recall": 0.9197651663405088, | |
| "eval_runtime": 30.8585, | |
| "eval_samples_per_second": 33.119, | |
| "eval_steps_per_second": 1.037, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.258547306060791, | |
| "learning_rate": 0.00018179542904605417, | |
| "loss": 0.4449, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8297455968688845, | |
| "eval_f1": 0.8406593406593407, | |
| "eval_loss": 0.4051818251609802, | |
| "eval_precision": 0.7900172117039587, | |
| "eval_recall": 0.898238747553816, | |
| "eval_runtime": 31.0429, | |
| "eval_samples_per_second": 32.922, | |
| "eval_steps_per_second": 1.031, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 5.823967933654785, | |
| "learning_rate": 9.089771452302708e-05, | |
| "loss": 0.4122, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8268101761252447, | |
| "eval_f1": 0.8350419384902142, | |
| "eval_loss": 0.42232340574264526, | |
| "eval_precision": 0.797153024911032, | |
| "eval_recall": 0.8767123287671232, | |
| "eval_runtime": 32.2891, | |
| "eval_samples_per_second": 31.652, | |
| "eval_steps_per_second": 0.991, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.855549335479736, | |
| "learning_rate": 0.0, | |
| "loss": 0.3933, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8356164383561644, | |
| "eval_f1": 0.8464351005484461, | |
| "eval_loss": 0.40576428174972534, | |
| "eval_precision": 0.7941680960548885, | |
| "eval_recall": 0.9060665362035225, | |
| "eval_runtime": 30.4475, | |
| "eval_samples_per_second": 33.566, | |
| "eval_steps_per_second": 1.051, | |
| "step": 384 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 384, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 942780789120.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.999132297568699, | |
| "learning_rate": 0.00036359085809210834, | |
| "num_train_epochs": 4, | |
| "temperature": 14 | |
| } | |
| } | |