{ "best_metric": 1.2424818277359009, "best_model_checkpoint": "/content/pubmedbert-abstract-cord19/checkpoint-5000", "epoch": 4.0, "global_step": 37788, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.53, "learning_rate": 2.4970000000000003e-05, "loss": 1.27, "step": 5000 }, { "epoch": 0.53, "eval_accuracy": 0.7236279677533664, "eval_loss": 1.2424818277359009, "eval_runtime": 183.46, "eval_samples_per_second": 44.206, "eval_steps_per_second": 2.764, "step": 5000 }, { "epoch": 1.06, "learning_rate": 4.9965e-05, "loss": 1.2634, "step": 10000 }, { "epoch": 1.06, "eval_accuracy": 0.7140570760001546, "eval_loss": 1.312286138534546, "eval_runtime": 183.3534, "eval_samples_per_second": 44.232, "eval_steps_per_second": 2.765, "step": 10000 }, { "epoch": 1.59, "learning_rate": 4.1024902835756445e-05, "loss": 1.3041, "step": 15000 }, { "epoch": 1.59, "eval_accuracy": 0.7072068247593172, "eval_loss": 1.3583375215530396, "eval_runtime": 183.3479, "eval_samples_per_second": 44.233, "eval_steps_per_second": 2.765, "step": 15000 }, { "epoch": 2.12, "learning_rate": 3.2033611630919824e-05, "loss": 1.3829, "step": 20000 }, { "epoch": 2.12, "eval_accuracy": 0.7120832508972437, "eval_loss": 1.3589657545089722, "eval_runtime": 183.3751, "eval_samples_per_second": 44.226, "eval_steps_per_second": 2.765, "step": 20000 }, { "epoch": 2.65, "learning_rate": 2.303692241255218e-05, "loss": 1.3069, "step": 25000 }, { "epoch": 2.65, "eval_accuracy": 0.7153998687867195, "eval_loss": 1.3505500555038452, "eval_runtime": 183.3915, "eval_samples_per_second": 44.222, "eval_steps_per_second": 2.765, "step": 25000 }, { "epoch": 3.18, "learning_rate": 1.4040233194184541e-05, "loss": 1.2921, "step": 30000 }, { "epoch": 3.18, "eval_accuracy": 0.7159658967235916, "eval_loss": 1.3447710275650024, "eval_runtime": 183.3443, "eval_samples_per_second": 44.234, "eval_steps_per_second": 2.765, "step": 30000 }, { "epoch": 3.7, "learning_rate": 5.0471426515042465e-06, "loss": 1.2731, "step": 35000 }, { "epoch": 3.7, "eval_accuracy": 0.7177905274023574, "eval_loss": 1.3374903202056885, "eval_runtime": 183.3182, "eval_samples_per_second": 44.24, "eval_steps_per_second": 2.766, "step": 35000 }, { "epoch": 4.0, "step": 37788, "total_flos": 1.3985984051532e+17, "train_loss": 1.2965481223991413, "train_runtime": 37723.3327, "train_samples_per_second": 16.027, "train_steps_per_second": 1.002 } ], "max_steps": 37788, "num_train_epochs": 4, "total_flos": 1.3985984051532e+17, "trial_name": null, "trial_params": null }