{ "best_metric": 2.009047269821167, "best_model_checkpoint": "output/obladaet/checkpoint-103", "epoch": 1.0, "global_step": 103, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05, "learning_rate": 0.0001364038084940016, "loss": 2.6831, "step": 5 }, { "epoch": 0.1, "learning_rate": 0.0001340337156353132, "loss": 2.4776, "step": 10 }, { "epoch": 0.15, "learning_rate": 0.00013014473739485363, "loss": 2.3448, "step": 15 }, { "epoch": 0.19, "learning_rate": 0.0001248271469925082, "loss": 2.2619, "step": 20 }, { "epoch": 0.24, "learning_rate": 0.00011820437942275874, "loss": 2.186, "step": 25 }, { "epoch": 0.29, "learning_rate": 0.00011043016620995614, "loss": 2.1301, "step": 30 }, { "epoch": 0.34, "learning_rate": 0.00010168496690295909, "loss": 2.0606, "step": 35 }, { "epoch": 0.39, "learning_rate": 9.21717801433564e-05, "loss": 2.0095, "step": 40 }, { "epoch": 0.44, "learning_rate": 8.211143154318805e-05, "loss": 2.0714, "step": 45 }, { "epoch": 0.49, "learning_rate": 7.173744775268251e-05, "loss": 2.0894, "step": 50 }, { "epoch": 0.53, "learning_rate": 6.12906357041245e-05, "loss": 2.1234, "step": 55 }, { "epoch": 0.58, "learning_rate": 5.101349286159054e-05, "loss": 2.0927, "step": 60 }, { "epoch": 0.63, "learning_rate": 4.114457822907933e-05, "loss": 2.0018, "step": 65 }, { "epoch": 0.68, "learning_rate": 3.1912974780463516e-05, "loss": 1.9459, "step": 70 }, { "epoch": 0.73, "learning_rate": 2.3532971852551605e-05, "loss": 1.9794, "step": 75 }, { "epoch": 0.78, "learning_rate": 1.6199090936639392e-05, "loss": 2.0329, "step": 80 }, { "epoch": 0.83, "learning_rate": 1.0081570332763395e-05, "loss": 1.9542, "step": 85 }, { "epoch": 0.87, "learning_rate": 5.322413479482709e-06, "loss": 2.0481, "step": 90 }, { "epoch": 0.92, "learning_rate": 2.032092687651423e-06, "loss": 2.02, "step": 95 }, { "epoch": 0.97, "learning_rate": 2.869847930266052e-07, "loss": 2.038, "step": 100 }, { "epoch": 1.0, "eval_loss": 2.009047269821167, "eval_runtime": 5.9275, "eval_samples_per_second": 22.775, "eval_steps_per_second": 2.868, "step": 103 } ], "max_steps": 103, "num_train_epochs": 1, "total_flos": 107129733120000.0, "trial_name": null, "trial_params": null }