{ "best_metric": 0.8567621097797157, "best_model_checkpoint": "./output//26_roberta-large_nace_5__5e-6_0.01_0.06_07-21-22_10-40/checkpoint-20000", "epoch": 0.09387489761768979, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 1.5645779550966127e-07, "loss": 0.4615, "step": 2000 }, { "epoch": 0.01, "eval_accuracy": 0.15318914123679223, "eval_f1": 0.27608921192458447, "eval_loss": 0.28353968262672424, "eval_roc_auc": 0.5800321310271234, "eval_runtime": 11156.151, "eval_samples_per_second": 33.95, "step": 2000 }, { "epoch": 0.02, "learning_rate": 3.1291559101932254e-07, "loss": 0.231, "step": 4000 }, { "epoch": 0.02, "eval_accuracy": 0.6693737887916695, "eval_f1": 0.7682238795441866, "eval_loss": 0.17183057963848114, "eval_roc_auc": 0.8419285809208303, "eval_runtime": 11148.4174, "eval_samples_per_second": 33.974, "step": 4000 }, { "epoch": 0.03, "learning_rate": 4.6937338652898386e-07, "loss": 0.1639, "step": 6000 }, { "epoch": 0.03, "eval_accuracy": 0.7266088278935668, "eval_f1": 0.8098187138286067, "eval_loss": 0.14115960896015167, "eval_roc_auc": 0.8728687872480946, "eval_runtime": 11147.6817, "eval_samples_per_second": 33.976, "step": 6000 }, { "epoch": 0.04, "learning_rate": 6.258311820386451e-07, "loss": 0.1414, "step": 8000 }, { "epoch": 0.04, "eval_accuracy": 0.7431657487445678, "eval_f1": 0.8278856538317427, "eval_loss": 0.12584036588668823, "eval_roc_auc": 0.8876332405370599, "eval_runtime": 11150.482, "eval_samples_per_second": 33.968, "step": 8000 }, { "epoch": 0.05, "learning_rate": 7.822889775483064e-07, "loss": 0.1283, "step": 10000 }, { "epoch": 0.05, "eval_accuracy": 0.7535894010360287, "eval_f1": 0.8360526710766673, "eval_loss": 0.11795546859502792, "eval_roc_auc": 0.8932732282506214, "eval_runtime": 11151.9405, "eval_samples_per_second": 33.963, "step": 10000 }, { "epoch": 0.06, "learning_rate": 9.387467730579677e-07, "loss": 0.1211, "step": 12000 }, { "epoch": 0.06, "eval_accuracy": 0.761145756876495, "eval_f1": 0.8441615840183859, "eval_loss": 0.11228282749652863, "eval_roc_auc": 0.9002726570618076, "eval_runtime": 11156.2585, "eval_samples_per_second": 33.95, "step": 12000 }, { "epoch": 0.07, "learning_rate": 1.095204568567629e-06, "loss": 0.1136, "step": 14000 }, { "epoch": 0.07, "eval_accuracy": 0.7673318301588894, "eval_f1": 0.8490507193175131, "eval_loss": 0.10872343182563782, "eval_roc_auc": 0.9023344325033343, "eval_runtime": 11157.3963, "eval_samples_per_second": 33.946, "step": 14000 }, { "epoch": 0.08, "learning_rate": 1.2516623640772901e-06, "loss": 0.1125, "step": 16000 }, { "epoch": 0.08, "eval_accuracy": 0.7712156175248315, "eval_f1": 0.8527881220679983, "eval_loss": 0.10447360575199127, "eval_roc_auc": 0.9056321731592804, "eval_runtime": 11156.4417, "eval_samples_per_second": 33.949, "step": 16000 }, { "epoch": 0.08, "learning_rate": 1.4081201595869516e-06, "loss": 0.1085, "step": 18000 }, { "epoch": 0.08, "eval_accuracy": 0.7746769671079381, "eval_f1": 0.8550148801059216, "eval_loss": 0.102951779961586, "eval_roc_auc": 0.9065747008981477, "eval_runtime": 11163.3791, "eval_samples_per_second": 33.928, "step": 18000 }, { "epoch": 0.09, "learning_rate": 1.5645779550966128e-06, "loss": 0.1063, "step": 20000 }, { "epoch": 0.09, "eval_accuracy": 0.7812854781731678, "eval_f1": 0.8567621097797157, "eval_loss": 0.10163237154483795, "eval_roc_auc": 0.904867851842597, "eval_runtime": 11164.6828, "eval_samples_per_second": 33.924, "step": 20000 } ], "max_steps": 1065245, "num_train_epochs": 5, "total_flos": 6.9868180733952e+16, "trial_name": null, "trial_params": null }