| { | |
| "best_metric": 0.4399268329143524, | |
| "best_model_checkpoint": "hBERTv2_new_pretrain_qqp/checkpoint-8529", | |
| "epoch": 8.0, | |
| "global_step": 22744, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3.9200000000000004e-05, | |
| "loss": 0.5303, | |
| "step": 2843 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7608211723967351, | |
| "eval_combined_score": 0.6928911943274398, | |
| "eval_f1": 0.6249612162581446, | |
| "eval_loss": 0.48931893706321716, | |
| "eval_runtime": 69.0476, | |
| "eval_samples_per_second": 585.538, | |
| "eval_steps_per_second": 4.577, | |
| "step": 2843 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.4677, | |
| "step": 5686 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7773188226564433, | |
| "eval_combined_score": 0.7302284114338071, | |
| "eval_f1": 0.6831380002111711, | |
| "eval_loss": 0.4780506491661072, | |
| "eval_runtime": 69.0619, | |
| "eval_samples_per_second": 585.417, | |
| "eval_steps_per_second": 4.576, | |
| "step": 5686 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.4229, | |
| "step": 8529 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7902547613158546, | |
| "eval_combined_score": 0.7519615243640321, | |
| "eval_f1": 0.7136682874122096, | |
| "eval_loss": 0.4399268329143524, | |
| "eval_runtime": 69.0436, | |
| "eval_samples_per_second": 585.572, | |
| "eval_steps_per_second": 4.577, | |
| "step": 8529 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.3712, | |
| "step": 11372 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8017808558001484, | |
| "eval_combined_score": 0.7590396774044639, | |
| "eval_f1": 0.7162984990087793, | |
| "eval_loss": 0.4425596296787262, | |
| "eval_runtime": 69.1036, | |
| "eval_samples_per_second": 585.064, | |
| "eval_steps_per_second": 4.573, | |
| "step": 11372 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.3268, | |
| "step": 14215 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8106851348008904, | |
| "eval_combined_score": 0.7727644478133877, | |
| "eval_f1": 0.7348437608258851, | |
| "eval_loss": 0.4514910578727722, | |
| "eval_runtime": 69.0616, | |
| "eval_samples_per_second": 585.419, | |
| "eval_steps_per_second": 4.576, | |
| "step": 14215 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.2925, | |
| "step": 17058 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8118971061093248, | |
| "eval_combined_score": 0.7672875750435039, | |
| "eval_f1": 0.722678043977683, | |
| "eval_loss": 0.5221311450004578, | |
| "eval_runtime": 69.0569, | |
| "eval_samples_per_second": 585.459, | |
| "eval_steps_per_second": 4.576, | |
| "step": 17058 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.44e-05, | |
| "loss": 0.2614, | |
| "step": 19901 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8057630472421469, | |
| "eval_combined_score": 0.7792239164658441, | |
| "eval_f1": 0.7526847856895412, | |
| "eval_loss": 0.45176824927330017, | |
| "eval_runtime": 69.1339, | |
| "eval_samples_per_second": 584.807, | |
| "eval_steps_per_second": 4.571, | |
| "step": 19901 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3.3600000000000004e-05, | |
| "loss": 0.2389, | |
| "step": 22744 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8134306208261192, | |
| "eval_combined_score": 0.7867833614195149, | |
| "eval_f1": 0.7601361020129106, | |
| "eval_loss": 0.5231410264968872, | |
| "eval_runtime": 69.0842, | |
| "eval_samples_per_second": 585.228, | |
| "eval_steps_per_second": 4.574, | |
| "step": 22744 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 22744, | |
| "total_flos": 4.25959360175276e+17, | |
| "train_loss": 0.3639708930574753, | |
| "train_runtime": 14995.3156, | |
| "train_samples_per_second": 1213.199, | |
| "train_steps_per_second": 9.48 | |
| } | |
| ], | |
| "max_steps": 142150, | |
| "num_train_epochs": 50, | |
| "total_flos": 4.25959360175276e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |