{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "global_step": 21050, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.12, "learning_rate": 1.952684085510689e-05, "loss": 0.3622, "step": 500 }, { "epoch": 0.24, "learning_rate": 1.9051781472684088e-05, "loss": 0.277, "step": 1000 }, { "epoch": 0.36, "learning_rate": 1.8576722090261282e-05, "loss": 0.2499, "step": 1500 }, { "epoch": 0.48, "learning_rate": 1.810166270783848e-05, "loss": 0.2266, "step": 2000 }, { "epoch": 0.59, "learning_rate": 1.7627553444180522e-05, "loss": 0.2035, "step": 2500 }, { "epoch": 0.71, "learning_rate": 1.715249406175772e-05, "loss": 0.2061, "step": 3000 }, { "epoch": 0.83, "learning_rate": 1.667743467933492e-05, "loss": 0.2024, "step": 3500 }, { "epoch": 0.95, "learning_rate": 1.6202375296912117e-05, "loss": 0.1887, "step": 4000 }, { "epoch": 1.07, "learning_rate": 1.5728266033254158e-05, "loss": 0.1641, "step": 4500 }, { "epoch": 1.19, "learning_rate": 1.5253206650831356e-05, "loss": 0.1314, "step": 5000 }, { "epoch": 1.31, "learning_rate": 1.4779097387173397e-05, "loss": 0.1363, "step": 5500 }, { "epoch": 1.43, "learning_rate": 1.4304038004750596e-05, "loss": 0.1261, "step": 6000 }, { "epoch": 1.54, "learning_rate": 1.382897862232779e-05, "loss": 0.1367, "step": 6500 }, { "epoch": 1.66, "learning_rate": 1.3353919239904989e-05, "loss": 0.1363, "step": 7000 }, { "epoch": 1.78, "learning_rate": 1.2878859857482187e-05, "loss": 0.1378, "step": 7500 }, { "epoch": 1.9, "learning_rate": 1.2403800475059384e-05, "loss": 0.1343, "step": 8000 }, { "epoch": 2.02, "learning_rate": 1.192874109263658e-05, "loss": 0.132, "step": 8500 }, { "epoch": 2.14, "learning_rate": 1.1453681710213777e-05, "loss": 0.0911, "step": 9000 }, { "epoch": 2.26, "learning_rate": 1.097957244655582e-05, "loss": 0.0806, "step": 9500 }, { "epoch": 2.38, "learning_rate": 1.0505463182897864e-05, "loss": 0.0852, "step": 10000 }, { "epoch": 2.49, "learning_rate": 1.003040380047506e-05, "loss": 0.0944, "step": 10500 }, { "epoch": 2.61, "learning_rate": 9.555344418052257e-06, "loss": 0.0936, "step": 11000 }, { "epoch": 2.73, "learning_rate": 9.080285035629454e-06, "loss": 0.0945, "step": 11500 }, { "epoch": 2.85, "learning_rate": 8.60522565320665e-06, "loss": 0.1064, "step": 12000 }, { "epoch": 2.97, "learning_rate": 8.130166270783849e-06, "loss": 0.0968, "step": 12500 }, { "epoch": 3.09, "learning_rate": 7.656057007125892e-06, "loss": 0.0597, "step": 13000 }, { "epoch": 3.21, "learning_rate": 7.180997624703089e-06, "loss": 0.05, "step": 13500 }, { "epoch": 3.33, "learning_rate": 6.705938242280286e-06, "loss": 0.0577, "step": 14000 }, { "epoch": 3.44, "learning_rate": 6.230878859857483e-06, "loss": 0.0589, "step": 14500 }, { "epoch": 3.56, "learning_rate": 5.75581947743468e-06, "loss": 0.0616, "step": 15000 }, { "epoch": 3.68, "learning_rate": 5.281710213776723e-06, "loss": 0.0633, "step": 15500 }, { "epoch": 3.8, "learning_rate": 4.806650831353919e-06, "loss": 0.0593, "step": 16000 }, { "epoch": 3.92, "learning_rate": 4.331591448931117e-06, "loss": 0.0698, "step": 16500 }, { "epoch": 4.04, "learning_rate": 3.856532066508314e-06, "loss": 0.0487, "step": 17000 }, { "epoch": 4.16, "learning_rate": 3.381472684085511e-06, "loss": 0.0321, "step": 17500 }, { "epoch": 4.28, "learning_rate": 2.9073634204275536e-06, "loss": 0.0384, "step": 18000 }, { "epoch": 4.39, "learning_rate": 2.4323040380047506e-06, "loss": 0.0324, "step": 18500 }, { "epoch": 4.51, "learning_rate": 1.957244655581948e-06, "loss": 0.0359, "step": 19000 }, { "epoch": 4.63, "learning_rate": 1.4821852731591448e-06, "loss": 0.0342, "step": 19500 }, { "epoch": 4.75, "learning_rate": 1.0080760095011877e-06, "loss": 0.0338, "step": 20000 }, { "epoch": 4.87, "learning_rate": 5.330166270783848e-07, "loss": 0.0334, "step": 20500 }, { "epoch": 4.99, "learning_rate": 5.7957244655581954e-08, "loss": 0.0384, "step": 21000 }, { "epoch": 5.0, "step": 21050, "total_flos": 2.21503330843008e+16, "train_loss": 0.111734983768146, "train_runtime": 1050.6373, "train_samples_per_second": 320.515, "train_steps_per_second": 20.035 } ], "max_steps": 21050, "num_train_epochs": 5, "total_flos": 2.21503330843008e+16, "trial_name": null, "trial_params": null }