{ "best_metric": 0.8761239051818848, "best_model_checkpoint": "./tst-translation-output/checkpoint-3000", "epoch": 5.465310459996964, "eval_steps": 1500, "global_step": 9000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.3, "learning_rate": 5e-05, "loss": 1.562, "step": 500 }, { "epoch": 0.61, "learning_rate": 4.956224829276834e-05, "loss": 1.118, "step": 1000 }, { "epoch": 0.91, "learning_rate": 4.912449658553668e-05, "loss": 1.014, "step": 1500 }, { "epoch": 0.91, "eval_bleu": 5.9697, "eval_gen_len": 18.2307, "eval_loss": 0.9561266303062439, "eval_runtime": 1131.8043, "eval_samples_per_second": 11.638, "eval_steps_per_second": 1.455, "step": 1500 }, { "epoch": 1.21, "learning_rate": 4.868674487830503e-05, "loss": 0.8557, "step": 2000 }, { "epoch": 1.52, "learning_rate": 4.824899317107337e-05, "loss": 0.7997, "step": 2500 }, { "epoch": 1.82, "learning_rate": 4.781124146384171e-05, "loss": 0.7888, "step": 3000 }, { "epoch": 1.82, "eval_bleu": 10.9388, "eval_gen_len": 17.4752, "eval_loss": 0.8761239051818848, "eval_runtime": 1004.0048, "eval_samples_per_second": 13.119, "eval_steps_per_second": 1.64, "step": 3000 }, { "epoch": 2.13, "learning_rate": 4.737348975661005e-05, "loss": 0.705, "step": 3500 }, { "epoch": 2.43, "learning_rate": 4.69357380493784e-05, "loss": 0.61, "step": 4000 }, { "epoch": 2.73, "learning_rate": 4.649798634214674e-05, "loss": 0.6166, "step": 4500 }, { "epoch": 2.73, "eval_bleu": 11.0165, "eval_gen_len": 17.1855, "eval_loss": 0.8778244256973267, "eval_runtime": 987.8488, "eval_samples_per_second": 13.334, "eval_steps_per_second": 1.667, "step": 4500 }, { "epoch": 3.04, "learning_rate": 4.606023463491508e-05, "loss": 0.5985, "step": 5000 }, { "epoch": 3.34, "learning_rate": 4.5622482927683416e-05, "loss": 0.4585, "step": 5500 }, { "epoch": 3.64, "learning_rate": 4.518473122045176e-05, "loss": 0.474, "step": 6000 }, { "epoch": 3.64, "eval_bleu": 11.8779, "eval_gen_len": 17.1838, "eval_loss": 0.9008126258850098, "eval_runtime": 981.3519, "eval_samples_per_second": 13.422, "eval_steps_per_second": 1.678, "step": 6000 }, { "epoch": 3.95, "learning_rate": 4.474697951322011e-05, "loss": 0.4858, "step": 6500 }, { "epoch": 4.25, "learning_rate": 4.430922780598845e-05, "loss": 0.3623, "step": 7000 }, { "epoch": 4.55, "learning_rate": 4.387147609875679e-05, "loss": 0.3559, "step": 7500 }, { "epoch": 4.55, "eval_bleu": 12.6024, "eval_gen_len": 17.3179, "eval_loss": 0.9503829479217529, "eval_runtime": 990.8364, "eval_samples_per_second": 13.294, "eval_steps_per_second": 1.662, "step": 7500 }, { "epoch": 4.86, "learning_rate": 4.3433724391525126e-05, "loss": 0.371, "step": 8000 }, { "epoch": 5.16, "learning_rate": 4.299597268429347e-05, "loss": 0.3053, "step": 8500 }, { "epoch": 5.47, "learning_rate": 4.255822097706181e-05, "loss": 0.2588, "step": 9000 }, { "epoch": 5.47, "eval_bleu": 13.4321, "eval_gen_len": 17.1999, "eval_loss": 0.9977866411209106, "eval_runtime": 986.567, "eval_samples_per_second": 13.351, "eval_steps_per_second": 1.669, "step": 9000 } ], "logging_steps": 500, "max_steps": 57610, "num_train_epochs": 35, "save_steps": 1500, "total_flos": 1.248266746331136e+18, "trial_name": null, "trial_params": null }