| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.977777777777778, | |
| "eval_steps": 500, | |
| "global_step": 630, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.769179709180454e-06, | |
| "loss": 1.3675, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.111518328538394e-05, | |
| "loss": 1.1968, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.307246990474186e-05, | |
| "loss": 1.2128, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.4461186861587421e-05, | |
| "loss": 1.2122, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.5538359418360907e-05, | |
| "loss": 1.1925, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.6418473480945343e-05, | |
| "loss": 1.1607, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 1.7162599318057647e-05, | |
| "loss": 1.1603, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.7807190437790904e-05, | |
| "loss": 1.1679, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1.8375760100303265e-05, | |
| "loss": 1.148, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.8884362994564394e-05, | |
| "loss": 1.1599, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.9344450276772258e-05, | |
| "loss": 1.178, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 1.9764477057148826e-05, | |
| "loss": 1.1472, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1571, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1796, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1968, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1665, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1686, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1733, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1516, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1518, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 2e-05, | |
| "loss": 1.162, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1181, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2e-05, | |
| "loss": 1.178, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1601, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1351, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2e-05, | |
| "loss": 0.898, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7484, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7723, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7332, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6949, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6972, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7155, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7311, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7117, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7201, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2e-05, | |
| "loss": 0.712, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7359, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7136, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7241, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7362, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7455, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7133, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7128, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7137, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7365, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7367, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2e-05, | |
| "loss": 0.735, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7126, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7302, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7651, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6033, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3221, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2935, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2851, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2795, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2897, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2749, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2825, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2744, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2799, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2773, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2849, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2811, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2895, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3122, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2963, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 2e-05, | |
| "loss": 0.29, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3038, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2975, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3048, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3017, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2956, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2958, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2e-05, | |
| "loss": 0.309, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3017, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2865, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0803, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0832, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0768, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0784, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0728, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0807, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0764, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0784, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0789, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0848, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0775, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0792, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0858, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0788, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0774, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0794, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0848, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0857, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0806, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0822, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0847, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0849, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0854, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0819, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0904, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0482, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0324, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0334, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 2e-05, | |
| "loss": 0.031, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0353, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0332, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0325, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0294, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0326, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0299, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0303, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0316, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0337, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0328, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0289, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0301, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0327, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0326, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0298, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0321, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0329, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0307, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0333, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0302, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0333, | |
| "step": 630 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 630, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "total_flos": 291660538544128.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |