| { | |
| "best_metric": 0.3039037883281708, | |
| "best_model_checkpoint": "hBERTv1_qqp/checkpoint-4266", | |
| "epoch": 8.0, | |
| "global_step": 11376, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.4011, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8285678951273806, | |
| "eval_combined_score": 0.8116115925137688, | |
| "eval_f1": 0.794655289900157, | |
| "eval_loss": 0.3665443956851959, | |
| "eval_runtime": 50.6077, | |
| "eval_samples_per_second": 798.89, | |
| "eval_steps_per_second": 3.122, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.3026, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8625278258718774, | |
| "eval_combined_score": 0.8398193588062908, | |
| "eval_f1": 0.8171108917407043, | |
| "eval_loss": 0.3111252784729004, | |
| "eval_runtime": 50.4355, | |
| "eval_samples_per_second": 801.618, | |
| "eval_steps_per_second": 3.133, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.2472, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8679940638139995, | |
| "eval_combined_score": 0.8450796349525497, | |
| "eval_f1": 0.8221652060910999, | |
| "eval_loss": 0.3039037883281708, | |
| "eval_runtime": 50.4592, | |
| "eval_samples_per_second": 801.241, | |
| "eval_steps_per_second": 3.131, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.1983, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8736581746228048, | |
| "eval_combined_score": 0.8531681858438971, | |
| "eval_f1": 0.8326781970649895, | |
| "eval_loss": 0.32322049140930176, | |
| "eval_runtime": 50.5899, | |
| "eval_samples_per_second": 799.171, | |
| "eval_steps_per_second": 3.123, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.500070323488045e-05, | |
| "loss": 0.157, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8717289141726441, | |
| "eval_combined_score": 0.8455726540359703, | |
| "eval_f1": 0.8194163938992965, | |
| "eval_loss": 0.3742031753063202, | |
| "eval_runtime": 50.4756, | |
| "eval_samples_per_second": 800.98, | |
| "eval_steps_per_second": 3.13, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.400070323488045e-05, | |
| "loss": 0.1251, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8716299777393025, | |
| "eval_combined_score": 0.8431165534906415, | |
| "eval_f1": 0.8146031292419805, | |
| "eval_loss": 0.40085598826408386, | |
| "eval_runtime": 50.493, | |
| "eval_samples_per_second": 800.705, | |
| "eval_steps_per_second": 3.129, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 4.30014064697609e-05, | |
| "loss": 0.1009, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8699480583724957, | |
| "eval_combined_score": 0.8499536553423827, | |
| "eval_f1": 0.8299592523122696, | |
| "eval_loss": 0.4470857083797455, | |
| "eval_runtime": 50.5119, | |
| "eval_samples_per_second": 800.405, | |
| "eval_steps_per_second": 3.128, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.200210970464135e-05, | |
| "loss": 0.0828, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8781103141231759, | |
| "eval_combined_score": 0.8567737171363892, | |
| "eval_f1": 0.8354371201496026, | |
| "eval_loss": 0.4176293909549713, | |
| "eval_runtime": 50.4947, | |
| "eval_samples_per_second": 800.677, | |
| "eval_steps_per_second": 3.129, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 11376, | |
| "total_flos": 3.676621197992264e+17, | |
| "train_loss": 0.20186867962071328, | |
| "train_runtime": 10978.3186, | |
| "train_samples_per_second": 1657.112, | |
| "train_steps_per_second": 6.476 | |
| } | |
| ], | |
| "max_steps": 71100, | |
| "num_train_epochs": 50, | |
| "total_flos": 3.676621197992264e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |