{ "best_metric": NaN, "best_model_checkpoint": "/scratch/czm5kz/finetuned_pythia70M_nondeduped_cp_14300032_2_0.0003_sequential/checkpoint-20", "epoch": 1.9950124688279303, "eval_steps": 20, "global_step": 2800, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 5 }, { "epoch": 0.01, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 10 }, { "epoch": 0.01, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 15 }, { "epoch": 0.01, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 20 }, { "epoch": 0.01, "eval_loss": NaN, "eval_runtime": 16.1863, "eval_samples_per_second": 693.736, "eval_steps_per_second": 86.74, "step": 20 }, { "epoch": 0.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 25 }, { "epoch": 0.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 30 }, { "epoch": 0.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 35 }, { "epoch": 0.03, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 40 }, { "epoch": 0.03, "eval_loss": NaN, "eval_runtime": 16.3333, "eval_samples_per_second": 687.492, "eval_steps_per_second": 85.959, "step": 40 }, { "epoch": 0.03, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 45 }, { "epoch": 0.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 50 }, { "epoch": 0.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 55 }, { "epoch": 0.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 60 }, { "epoch": 0.04, "eval_loss": NaN, "eval_runtime": 16.2391, "eval_samples_per_second": 691.481, "eval_steps_per_second": 86.458, "step": 60 }, { "epoch": 0.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 65 }, { "epoch": 0.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 70 }, { "epoch": 0.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 75 }, { "epoch": 0.06, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 80 }, { "epoch": 0.06, "eval_loss": NaN, "eval_runtime": 16.018, "eval_samples_per_second": 701.023, "eval_steps_per_second": 87.651, "step": 80 }, { "epoch": 0.06, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 85 }, { "epoch": 0.06, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 90 }, { "epoch": 0.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 95 }, { "epoch": 0.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 100 }, { "epoch": 0.07, "eval_loss": NaN, "eval_runtime": 16.0813, "eval_samples_per_second": 698.263, "eval_steps_per_second": 87.306, "step": 100 }, { "epoch": 0.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 105 }, { "epoch": 0.08, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 110 }, { "epoch": 0.08, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 115 }, { "epoch": 0.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 120 }, { "epoch": 0.09, "eval_loss": NaN, "eval_runtime": 16.2065, "eval_samples_per_second": 692.869, "eval_steps_per_second": 86.632, "step": 120 }, { "epoch": 0.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 125 }, { "epoch": 0.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 130 }, { "epoch": 0.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 135 }, { "epoch": 0.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 140 }, { "epoch": 0.1, "eval_loss": NaN, "eval_runtime": 15.9842, "eval_samples_per_second": 702.505, "eval_steps_per_second": 87.837, "step": 140 }, { "epoch": 0.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 145 }, { "epoch": 0.11, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 150 }, { "epoch": 0.11, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 155 }, { "epoch": 0.11, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 160 }, { "epoch": 0.11, "eval_loss": NaN, "eval_runtime": 16.1805, "eval_samples_per_second": 693.984, "eval_steps_per_second": 86.771, "step": 160 }, { "epoch": 0.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 165 }, { "epoch": 0.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.3158, "step": 170 }, { "epoch": 0.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 175 }, { "epoch": 0.13, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 180 }, { "epoch": 0.13, "eval_loss": NaN, "eval_runtime": 16.165, "eval_samples_per_second": 694.649, "eval_steps_per_second": 86.854, "step": 180 }, { "epoch": 0.13, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 185 }, { "epoch": 0.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 190 }, { "epoch": 0.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 195 }, { "epoch": 0.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 200 }, { "epoch": 0.14, "eval_loss": NaN, "eval_runtime": 15.9948, "eval_samples_per_second": 702.04, "eval_steps_per_second": 87.778, "step": 200 }, { "epoch": 0.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 205 }, { "epoch": 0.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 210 }, { "epoch": 0.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 215 }, { "epoch": 0.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 220 }, { "epoch": 0.16, "eval_loss": NaN, "eval_runtime": 16.1728, "eval_samples_per_second": 694.313, "eval_steps_per_second": 86.812, "step": 220 }, { "epoch": 0.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 225 }, { "epoch": 0.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 230 }, { "epoch": 0.17, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 235 }, { "epoch": 0.17, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 240 }, { "epoch": 0.17, "eval_loss": NaN, "eval_runtime": 16.2473, "eval_samples_per_second": 691.13, "eval_steps_per_second": 86.414, "step": 240 }, { "epoch": 0.17, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 107.6567, "step": 245 }, { "epoch": 0.18, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 250 }, { "epoch": 0.18, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 255 }, { "epoch": 0.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 260 }, { "epoch": 0.19, "eval_loss": NaN, "eval_runtime": 16.1347, "eval_samples_per_second": 695.952, "eval_steps_per_second": 87.017, "step": 260 }, { "epoch": 0.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 265 }, { "epoch": 0.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 270 }, { "epoch": 0.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 275 }, { "epoch": 0.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 280 }, { "epoch": 0.2, "eval_loss": NaN, "eval_runtime": 16.2493, "eval_samples_per_second": 691.044, "eval_steps_per_second": 86.404, "step": 280 }, { "epoch": 0.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 285 }, { "epoch": 0.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 290 }, { "epoch": 0.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 295 }, { "epoch": 0.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 300 }, { "epoch": 0.21, "eval_loss": NaN, "eval_runtime": 16.1624, "eval_samples_per_second": 694.759, "eval_steps_per_second": 86.868, "step": 300 }, { "epoch": 0.22, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 364.1836, "step": 305 }, { "epoch": 0.22, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 310 }, { "epoch": 0.22, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 315 }, { "epoch": 0.23, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 320 }, { "epoch": 0.23, "eval_loss": NaN, "eval_runtime": 15.9725, "eval_samples_per_second": 703.022, "eval_steps_per_second": 87.901, "step": 320 }, { "epoch": 0.23, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 325 }, { "epoch": 0.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 330 }, { "epoch": 0.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 335 }, { "epoch": 0.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 340 }, { "epoch": 0.24, "eval_loss": NaN, "eval_runtime": 16.1818, "eval_samples_per_second": 693.927, "eval_steps_per_second": 86.764, "step": 340 }, { "epoch": 0.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 345 }, { "epoch": 0.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 350 }, { "epoch": 0.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 355 }, { "epoch": 0.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 360 }, { "epoch": 0.26, "eval_loss": NaN, "eval_runtime": 15.9768, "eval_samples_per_second": 702.833, "eval_steps_per_second": 87.878, "step": 360 }, { "epoch": 0.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 365 }, { "epoch": 0.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 370 }, { "epoch": 0.27, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 375 }, { "epoch": 0.27, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 380 }, { "epoch": 0.27, "eval_loss": NaN, "eval_runtime": 15.9923, "eval_samples_per_second": 702.152, "eval_steps_per_second": 87.793, "step": 380 }, { "epoch": 0.27, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 385 }, { "epoch": 0.28, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 390 }, { "epoch": 0.28, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 395 }, { "epoch": 0.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 400 }, { "epoch": 0.29, "eval_loss": NaN, "eval_runtime": 16.3253, "eval_samples_per_second": 687.828, "eval_steps_per_second": 86.001, "step": 400 }, { "epoch": 0.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 405 }, { "epoch": 0.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 410 }, { "epoch": 0.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 415 }, { "epoch": 0.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 420 }, { "epoch": 0.3, "eval_loss": NaN, "eval_runtime": 16.1692, "eval_samples_per_second": 694.47, "eval_steps_per_second": 86.832, "step": 420 }, { "epoch": 0.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 425 }, { "epoch": 0.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 430 }, { "epoch": 0.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 435 }, { "epoch": 0.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 440 }, { "epoch": 0.31, "eval_loss": NaN, "eval_runtime": 15.9356, "eval_samples_per_second": 704.65, "eval_steps_per_second": 88.105, "step": 440 }, { "epoch": 0.32, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 445 }, { "epoch": 0.32, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 450 }, { "epoch": 0.32, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 455 }, { "epoch": 0.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 460 }, { "epoch": 0.33, "eval_loss": NaN, "eval_runtime": 16.1082, "eval_samples_per_second": 697.099, "eval_steps_per_second": 87.161, "step": 460 }, { "epoch": 0.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 465 }, { "epoch": 0.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 470 }, { "epoch": 0.34, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 475 }, { "epoch": 0.34, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 480 }, { "epoch": 0.34, "eval_loss": NaN, "eval_runtime": 15.9065, "eval_samples_per_second": 705.938, "eval_steps_per_second": 88.266, "step": 480 }, { "epoch": 0.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 485 }, { "epoch": 0.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 490 }, { "epoch": 0.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 495 }, { "epoch": 0.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 500 }, { "epoch": 0.36, "eval_loss": NaN, "eval_runtime": 16.1844, "eval_samples_per_second": 693.815, "eval_steps_per_second": 86.75, "step": 500 }, { "epoch": 0.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 505 }, { "epoch": 0.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 510 }, { "epoch": 0.37, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 515 }, { "epoch": 0.37, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 851.5, "step": 520 }, { "epoch": 0.37, "eval_loss": NaN, "eval_runtime": 16.1417, "eval_samples_per_second": 695.653, "eval_steps_per_second": 86.98, "step": 520 }, { "epoch": 0.37, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 525 }, { "epoch": 0.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 530 }, { "epoch": 0.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 535 }, { "epoch": 0.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 540 }, { "epoch": 0.38, "eval_loss": NaN, "eval_runtime": 15.9976, "eval_samples_per_second": 701.919, "eval_steps_per_second": 87.763, "step": 540 }, { "epoch": 0.39, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 545 }, { "epoch": 0.39, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 550 }, { "epoch": 0.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 555 }, { "epoch": 0.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 560 }, { "epoch": 0.4, "eval_loss": NaN, "eval_runtime": 16.1555, "eval_samples_per_second": 695.055, "eval_steps_per_second": 86.905, "step": 560 }, { "epoch": 0.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 565 }, { "epoch": 0.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 570 }, { "epoch": 0.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 575 }, { "epoch": 0.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 580 }, { "epoch": 0.41, "eval_loss": NaN, "eval_runtime": 16.1891, "eval_samples_per_second": 693.614, "eval_steps_per_second": 86.725, "step": 580 }, { "epoch": 0.42, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 585 }, { "epoch": 0.42, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 590 }, { "epoch": 0.42, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 595 }, { "epoch": 0.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 600 }, { "epoch": 0.43, "eval_loss": NaN, "eval_runtime": 16.034, "eval_samples_per_second": 700.323, "eval_steps_per_second": 87.564, "step": 600 }, { "epoch": 0.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 605 }, { "epoch": 0.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 610 }, { "epoch": 0.44, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 615 }, { "epoch": 0.44, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 620 }, { "epoch": 0.44, "eval_loss": NaN, "eval_runtime": 16.1594, "eval_samples_per_second": 694.889, "eval_steps_per_second": 86.884, "step": 620 }, { "epoch": 0.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 625 }, { "epoch": 0.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 630 }, { "epoch": 0.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 635 }, { "epoch": 0.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 640 }, { "epoch": 0.46, "eval_loss": NaN, "eval_runtime": 16.1792, "eval_samples_per_second": 694.038, "eval_steps_per_second": 86.778, "step": 640 }, { "epoch": 0.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 645 }, { "epoch": 0.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 650 }, { "epoch": 0.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 655 }, { "epoch": 0.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 4.2227, "step": 660 }, { "epoch": 0.47, "eval_loss": NaN, "eval_runtime": 16.0542, "eval_samples_per_second": 699.443, "eval_steps_per_second": 87.454, "step": 660 }, { "epoch": 0.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 665 }, { "epoch": 0.48, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 670 }, { "epoch": 0.48, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 675 }, { "epoch": 0.48, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 680 }, { "epoch": 0.48, "eval_loss": NaN, "eval_runtime": 16.311, "eval_samples_per_second": 688.43, "eval_steps_per_second": 86.077, "step": 680 }, { "epoch": 0.49, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 685 }, { "epoch": 0.49, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 690 }, { "epoch": 0.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 695 }, { "epoch": 0.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 700 }, { "epoch": 0.5, "eval_loss": NaN, "eval_runtime": 16.1193, "eval_samples_per_second": 696.618, "eval_steps_per_second": 87.1, "step": 700 }, { "epoch": 0.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 705 }, { "epoch": 0.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 710 }, { "epoch": 0.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 715 }, { "epoch": 0.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 720 }, { "epoch": 0.51, "eval_loss": NaN, "eval_runtime": 15.972, "eval_samples_per_second": 703.041, "eval_steps_per_second": 87.904, "step": 720 }, { "epoch": 0.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 725 }, { "epoch": 0.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 730 }, { "epoch": 0.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 735 }, { "epoch": 0.53, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 740 }, { "epoch": 0.53, "eval_loss": NaN, "eval_runtime": 16.2427, "eval_samples_per_second": 691.327, "eval_steps_per_second": 86.439, "step": 740 }, { "epoch": 0.53, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 745 }, { "epoch": 0.53, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 750 }, { "epoch": 0.54, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 755 }, { "epoch": 0.54, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 760 }, { "epoch": 0.54, "eval_loss": NaN, "eval_runtime": 16.1102, "eval_samples_per_second": 697.012, "eval_steps_per_second": 87.15, "step": 760 }, { "epoch": 0.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 765 }, { "epoch": 0.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 770 }, { "epoch": 0.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 775 }, { "epoch": 0.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 780 }, { "epoch": 0.56, "eval_loss": NaN, "eval_runtime": 16.1434, "eval_samples_per_second": 695.581, "eval_steps_per_second": 86.971, "step": 780 }, { "epoch": 0.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 785 }, { "epoch": 0.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 790 }, { "epoch": 0.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 795 }, { "epoch": 0.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 800 }, { "epoch": 0.57, "eval_loss": NaN, "eval_runtime": 16.286, "eval_samples_per_second": 689.488, "eval_steps_per_second": 86.209, "step": 800 }, { "epoch": 0.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 805 }, { "epoch": 0.58, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 810 }, { "epoch": 0.58, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 815 }, { "epoch": 0.58, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 820 }, { "epoch": 0.58, "eval_loss": NaN, "eval_runtime": 16.0438, "eval_samples_per_second": 699.898, "eval_steps_per_second": 87.511, "step": 820 }, { "epoch": 0.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 3.1509, "step": 825 }, { "epoch": 0.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 830 }, { "epoch": 0.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 835 }, { "epoch": 0.6, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.9594, "step": 840 }, { "epoch": 0.6, "eval_loss": NaN, "eval_runtime": 16.2296, "eval_samples_per_second": 691.886, "eval_steps_per_second": 86.509, "step": 840 }, { "epoch": 0.6, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 845 }, { "epoch": 0.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 850 }, { "epoch": 0.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 855 }, { "epoch": 0.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 860 }, { "epoch": 0.61, "eval_loss": NaN, "eval_runtime": 16.1902, "eval_samples_per_second": 693.568, "eval_steps_per_second": 86.719, "step": 860 }, { "epoch": 0.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 1.0598, "step": 865 }, { "epoch": 0.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 870 }, { "epoch": 0.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 330.1973, "step": 875 }, { "epoch": 0.63, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 880 }, { "epoch": 0.63, "eval_loss": NaN, "eval_runtime": 16.0644, "eval_samples_per_second": 699.001, "eval_steps_per_second": 87.398, "step": 880 }, { "epoch": 0.63, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 885 }, { "epoch": 0.63, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 890 }, { "epoch": 0.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 895 }, { "epoch": 0.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 900 }, { "epoch": 0.64, "eval_loss": NaN, "eval_runtime": 16.254, "eval_samples_per_second": 690.844, "eval_steps_per_second": 86.379, "step": 900 }, { "epoch": 0.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 905 }, { "epoch": 0.65, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 910 }, { "epoch": 0.65, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 915 }, { "epoch": 0.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 920 }, { "epoch": 0.66, "eval_loss": NaN, "eval_runtime": 16.2774, "eval_samples_per_second": 689.852, "eval_steps_per_second": 86.255, "step": 920 }, { "epoch": 0.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 925 }, { "epoch": 0.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 930 }, { "epoch": 0.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 935 }, { "epoch": 0.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 940 }, { "epoch": 0.67, "eval_loss": NaN, "eval_runtime": 15.9234, "eval_samples_per_second": 705.188, "eval_steps_per_second": 88.172, "step": 940 }, { "epoch": 0.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 945 }, { "epoch": 0.68, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 950 }, { "epoch": 0.68, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 955 }, { "epoch": 0.68, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 960 }, { "epoch": 0.68, "eval_loss": NaN, "eval_runtime": 16.1163, "eval_samples_per_second": 696.749, "eval_steps_per_second": 87.117, "step": 960 }, { "epoch": 0.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 965 }, { "epoch": 0.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 970 }, { "epoch": 0.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 975 }, { "epoch": 0.7, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 980 }, { "epoch": 0.7, "eval_loss": NaN, "eval_runtime": 16.1208, "eval_samples_per_second": 696.552, "eval_steps_per_second": 87.092, "step": 980 }, { "epoch": 0.7, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 985 }, { "epoch": 0.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 67.9312, "step": 990 }, { "epoch": 0.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 995 }, { "epoch": 0.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1000 }, { "epoch": 0.71, "eval_loss": NaN, "eval_runtime": 15.9291, "eval_samples_per_second": 704.938, "eval_steps_per_second": 88.141, "step": 1000 }, { "epoch": 0.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1005 }, { "epoch": 0.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1010 }, { "epoch": 0.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1015 }, { "epoch": 0.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1020 }, { "epoch": 0.73, "eval_loss": NaN, "eval_runtime": 15.9195, "eval_samples_per_second": 705.363, "eval_steps_per_second": 88.194, "step": 1020 }, { "epoch": 0.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1025 }, { "epoch": 0.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1030 }, { "epoch": 0.74, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1035 }, { "epoch": 0.74, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1040 }, { "epoch": 0.74, "eval_loss": NaN, "eval_runtime": 16.0887, "eval_samples_per_second": 697.944, "eval_steps_per_second": 87.266, "step": 1040 }, { "epoch": 0.74, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1045 }, { "epoch": 0.75, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1050 }, { "epoch": 0.75, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1055 }, { "epoch": 0.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1060 }, { "epoch": 0.76, "eval_loss": NaN, "eval_runtime": 15.9033, "eval_samples_per_second": 706.08, "eval_steps_per_second": 88.284, "step": 1060 }, { "epoch": 0.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1065 }, { "epoch": 0.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1070 }, { "epoch": 0.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1075 }, { "epoch": 0.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1080 }, { "epoch": 0.77, "eval_loss": NaN, "eval_runtime": 15.8875, "eval_samples_per_second": 706.78, "eval_steps_per_second": 88.371, "step": 1080 }, { "epoch": 0.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1085 }, { "epoch": 0.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1090 }, { "epoch": 0.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1095 }, { "epoch": 0.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1100 }, { "epoch": 0.78, "eval_loss": NaN, "eval_runtime": 16.0767, "eval_samples_per_second": 698.465, "eval_steps_per_second": 87.331, "step": 1100 }, { "epoch": 0.79, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1105 }, { "epoch": 0.79, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1110 }, { "epoch": 0.79, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1115 }, { "epoch": 0.8, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1120 }, { "epoch": 0.8, "eval_loss": NaN, "eval_runtime": 15.9083, "eval_samples_per_second": 705.858, "eval_steps_per_second": 88.256, "step": 1120 }, { "epoch": 0.8, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1125 }, { "epoch": 0.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1130 }, { "epoch": 0.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1135 }, { "epoch": 0.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1140 }, { "epoch": 0.81, "eval_loss": NaN, "eval_runtime": 16.1038, "eval_samples_per_second": 697.289, "eval_steps_per_second": 87.184, "step": 1140 }, { "epoch": 0.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1145 }, { "epoch": 0.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1150 }, { "epoch": 0.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1155 }, { "epoch": 0.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1160 }, { "epoch": 0.83, "eval_loss": NaN, "eval_runtime": 16.0802, "eval_samples_per_second": 698.312, "eval_steps_per_second": 87.312, "step": 1160 }, { "epoch": 0.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 14.6949, "step": 1165 }, { "epoch": 0.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1170 }, { "epoch": 0.84, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1175 }, { "epoch": 0.84, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1180 }, { "epoch": 0.84, "eval_loss": NaN, "eval_runtime": 15.952, "eval_samples_per_second": 703.923, "eval_steps_per_second": 88.014, "step": 1180 }, { "epoch": 0.84, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1185 }, { "epoch": 0.85, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1190 }, { "epoch": 0.85, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1195 }, { "epoch": 0.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1200 }, { "epoch": 0.86, "eval_loss": NaN, "eval_runtime": 16.1511, "eval_samples_per_second": 695.248, "eval_steps_per_second": 86.929, "step": 1200 }, { "epoch": 0.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1205 }, { "epoch": 0.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1210 }, { "epoch": 0.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1215 }, { "epoch": 0.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1220 }, { "epoch": 0.87, "eval_loss": NaN, "eval_runtime": 16.1363, "eval_samples_per_second": 695.883, "eval_steps_per_second": 87.009, "step": 1220 }, { "epoch": 0.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1225 }, { "epoch": 0.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1230 }, { "epoch": 0.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1235 }, { "epoch": 0.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1240 }, { "epoch": 0.88, "eval_loss": NaN, "eval_runtime": 15.927, "eval_samples_per_second": 705.03, "eval_steps_per_second": 88.152, "step": 1240 }, { "epoch": 0.89, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1245 }, { "epoch": 0.89, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1250 }, { "epoch": 0.89, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1255 }, { "epoch": 0.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1260 }, { "epoch": 0.9, "eval_loss": NaN, "eval_runtime": 16.103, "eval_samples_per_second": 697.323, "eval_steps_per_second": 87.189, "step": 1260 }, { "epoch": 0.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 5.7933, "step": 1265 }, { "epoch": 0.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1270 }, { "epoch": 0.91, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1275 }, { "epoch": 0.91, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1280 }, { "epoch": 0.91, "eval_loss": NaN, "eval_runtime": 16.0834, "eval_samples_per_second": 698.171, "eval_steps_per_second": 87.295, "step": 1280 }, { "epoch": 0.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1285 }, { "epoch": 0.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1290 }, { "epoch": 0.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1295 }, { "epoch": 0.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1300 }, { "epoch": 0.93, "eval_loss": NaN, "eval_runtime": 15.9136, "eval_samples_per_second": 705.624, "eval_steps_per_second": 88.227, "step": 1300 }, { "epoch": 0.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1305 }, { "epoch": 0.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1310 }, { "epoch": 0.94, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1315 }, { "epoch": 0.94, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1320 }, { "epoch": 0.94, "eval_loss": NaN, "eval_runtime": 16.0934, "eval_samples_per_second": 697.739, "eval_steps_per_second": 87.241, "step": 1320 }, { "epoch": 0.94, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1325 }, { "epoch": 0.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1330 }, { "epoch": 0.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1335 }, { "epoch": 0.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1340 }, { "epoch": 0.95, "eval_loss": NaN, "eval_runtime": 15.9327, "eval_samples_per_second": 704.779, "eval_steps_per_second": 88.121, "step": 1340 }, { "epoch": 0.96, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1345 }, { "epoch": 0.96, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1350 }, { "epoch": 0.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1355 }, { "epoch": 0.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1360 }, { "epoch": 0.97, "eval_loss": NaN, "eval_runtime": 15.9525, "eval_samples_per_second": 703.904, "eval_steps_per_second": 88.011, "step": 1360 }, { "epoch": 0.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1365 }, { "epoch": 0.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1370 }, { "epoch": 0.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1375 }, { "epoch": 0.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1380 }, { "epoch": 0.98, "eval_loss": NaN, "eval_runtime": 16.1341, "eval_samples_per_second": 695.979, "eval_steps_per_second": 87.021, "step": 1380 }, { "epoch": 0.99, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1385 }, { "epoch": 0.99, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1390 }, { "epoch": 0.99, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1395 }, { "epoch": 1.0, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1400 }, { "epoch": 1.0, "eval_loss": NaN, "eval_runtime": 15.9261, "eval_samples_per_second": 705.071, "eval_steps_per_second": 88.157, "step": 1400 }, { "epoch": 1.0, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1405 }, { "epoch": 1.0, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1410 }, { "epoch": 1.01, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1415 }, { "epoch": 1.01, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1420 }, { "epoch": 1.01, "eval_loss": NaN, "eval_runtime": 16.1185, "eval_samples_per_second": 696.651, "eval_steps_per_second": 87.105, "step": 1420 }, { "epoch": 1.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1425 }, { "epoch": 1.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1430 }, { "epoch": 1.02, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1435 }, { "epoch": 1.03, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1440 }, { "epoch": 1.03, "eval_loss": NaN, "eval_runtime": 16.1075, "eval_samples_per_second": 697.131, "eval_steps_per_second": 87.165, "step": 1440 }, { "epoch": 1.03, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1445 }, { "epoch": 1.03, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1450 }, { "epoch": 1.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1455 }, { "epoch": 1.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1460 }, { "epoch": 1.04, "eval_loss": NaN, "eval_runtime": 15.9284, "eval_samples_per_second": 704.969, "eval_steps_per_second": 88.145, "step": 1460 }, { "epoch": 1.04, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1465 }, { "epoch": 1.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1470 }, { "epoch": 1.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1475 }, { "epoch": 1.05, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1480 }, { "epoch": 1.05, "eval_loss": NaN, "eval_runtime": 16.1079, "eval_samples_per_second": 697.111, "eval_steps_per_second": 87.162, "step": 1480 }, { "epoch": 1.06, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1485 }, { "epoch": 1.06, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1490 }, { "epoch": 1.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1495 }, { "epoch": 1.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1500 }, { "epoch": 1.07, "eval_loss": NaN, "eval_runtime": 16.1102, "eval_samples_per_second": 697.011, "eval_steps_per_second": 87.15, "step": 1500 }, { "epoch": 1.07, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1505 }, { "epoch": 1.08, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1510 }, { "epoch": 1.08, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1515 }, { "epoch": 1.08, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1520 }, { "epoch": 1.08, "eval_loss": NaN, "eval_runtime": 15.8974, "eval_samples_per_second": 706.342, "eval_steps_per_second": 88.316, "step": 1520 }, { "epoch": 1.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1525 }, { "epoch": 1.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1530 }, { "epoch": 1.09, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1535 }, { "epoch": 1.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1540 }, { "epoch": 1.1, "eval_loss": NaN, "eval_runtime": 16.0835, "eval_samples_per_second": 698.167, "eval_steps_per_second": 87.294, "step": 1540 }, { "epoch": 1.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1545 }, { "epoch": 1.1, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1550 }, { "epoch": 1.11, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1555 }, { "epoch": 1.11, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1560 }, { "epoch": 1.11, "eval_loss": NaN, "eval_runtime": 16.1034, "eval_samples_per_second": 697.305, "eval_steps_per_second": 87.186, "step": 1560 }, { "epoch": 1.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1565 }, { "epoch": 1.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1570 }, { "epoch": 1.12, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1575 }, { "epoch": 1.13, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1580 }, { "epoch": 1.13, "eval_loss": NaN, "eval_runtime": 15.9233, "eval_samples_per_second": 705.194, "eval_steps_per_second": 88.173, "step": 1580 }, { "epoch": 1.13, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1585 }, { "epoch": 1.13, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1590 }, { "epoch": 1.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1595 }, { "epoch": 1.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1600 }, { "epoch": 1.14, "eval_loss": NaN, "eval_runtime": 16.1179, "eval_samples_per_second": 696.681, "eval_steps_per_second": 87.108, "step": 1600 }, { "epoch": 1.14, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1605 }, { "epoch": 1.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1610 }, { "epoch": 1.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1615 }, { "epoch": 1.15, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1620 }, { "epoch": 1.15, "eval_loss": NaN, "eval_runtime": 15.9435, "eval_samples_per_second": 704.3, "eval_steps_per_second": 88.061, "step": 1620 }, { "epoch": 1.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1625 }, { "epoch": 1.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1630 }, { "epoch": 1.16, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1635 }, { "epoch": 1.17, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1640 }, { "epoch": 1.17, "eval_loss": NaN, "eval_runtime": 15.9139, "eval_samples_per_second": 705.608, "eval_steps_per_second": 88.225, "step": 1640 }, { "epoch": 1.17, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1645 }, { "epoch": 1.18, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.9002, "step": 1650 }, { "epoch": 1.18, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1655 }, { "epoch": 1.18, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1660 }, { "epoch": 1.18, "eval_loss": NaN, "eval_runtime": 16.1378, "eval_samples_per_second": 695.822, "eval_steps_per_second": 87.001, "step": 1660 }, { "epoch": 1.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1665 }, { "epoch": 1.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1670 }, { "epoch": 1.19, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1675 }, { "epoch": 1.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1680 }, { "epoch": 1.2, "eval_loss": NaN, "eval_runtime": 15.9444, "eval_samples_per_second": 704.259, "eval_steps_per_second": 88.056, "step": 1680 }, { "epoch": 1.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1685 }, { "epoch": 1.2, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1690 }, { "epoch": 1.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1695 }, { "epoch": 1.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1700 }, { "epoch": 1.21, "eval_loss": NaN, "eval_runtime": 15.9454, "eval_samples_per_second": 704.218, "eval_steps_per_second": 88.051, "step": 1700 }, { "epoch": 1.21, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1705 }, { "epoch": 1.22, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1710 }, { "epoch": 1.22, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1715 }, { "epoch": 1.23, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1720 }, { "epoch": 1.23, "eval_loss": NaN, "eval_runtime": 16.1205, "eval_samples_per_second": 696.565, "eval_steps_per_second": 87.094, "step": 1720 }, { "epoch": 1.23, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1725 }, { "epoch": 1.23, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1730 }, { "epoch": 1.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1735 }, { "epoch": 1.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1740 }, { "epoch": 1.24, "eval_loss": NaN, "eval_runtime": 15.9648, "eval_samples_per_second": 703.358, "eval_steps_per_second": 87.943, "step": 1740 }, { "epoch": 1.24, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1745 }, { "epoch": 1.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1750 }, { "epoch": 1.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1755 }, { "epoch": 1.25, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1760 }, { "epoch": 1.25, "eval_loss": NaN, "eval_runtime": 16.1129, "eval_samples_per_second": 696.897, "eval_steps_per_second": 87.135, "step": 1760 }, { "epoch": 1.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1765 }, { "epoch": 1.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1770 }, { "epoch": 1.26, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1775 }, { "epoch": 1.27, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1780 }, { "epoch": 1.27, "eval_loss": NaN, "eval_runtime": 16.1011, "eval_samples_per_second": 697.408, "eval_steps_per_second": 87.199, "step": 1780 }, { "epoch": 1.27, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1785 }, { "epoch": 1.28, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1790 }, { "epoch": 1.28, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1795 }, { "epoch": 1.28, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1800 }, { "epoch": 1.28, "eval_loss": NaN, "eval_runtime": 15.9086, "eval_samples_per_second": 705.846, "eval_steps_per_second": 88.254, "step": 1800 }, { "epoch": 1.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1805 }, { "epoch": 1.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1810 }, { "epoch": 1.29, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1815 }, { "epoch": 1.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1820 }, { "epoch": 1.3, "eval_loss": NaN, "eval_runtime": 16.0953, "eval_samples_per_second": 697.659, "eval_steps_per_second": 87.231, "step": 1820 }, { "epoch": 1.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1825 }, { "epoch": 1.3, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1830 }, { "epoch": 1.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1835 }, { "epoch": 1.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1840 }, { "epoch": 1.31, "eval_loss": NaN, "eval_runtime": 16.1084, "eval_samples_per_second": 697.09, "eval_steps_per_second": 87.16, "step": 1840 }, { "epoch": 1.31, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1845 }, { "epoch": 1.32, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1850 }, { "epoch": 1.32, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1855 }, { "epoch": 1.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1860 }, { "epoch": 1.33, "eval_loss": NaN, "eval_runtime": 15.9291, "eval_samples_per_second": 704.937, "eval_steps_per_second": 88.141, "step": 1860 }, { "epoch": 1.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1865 }, { "epoch": 1.33, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1870 }, { "epoch": 1.34, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1875 }, { "epoch": 1.34, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1880 }, { "epoch": 1.34, "eval_loss": NaN, "eval_runtime": 16.1362, "eval_samples_per_second": 695.887, "eval_steps_per_second": 87.009, "step": 1880 }, { "epoch": 1.34, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1885 }, { "epoch": 1.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1890 }, { "epoch": 1.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1895 }, { "epoch": 1.35, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1900 }, { "epoch": 1.35, "eval_loss": NaN, "eval_runtime": 16.0979, "eval_samples_per_second": 697.544, "eval_steps_per_second": 87.216, "step": 1900 }, { "epoch": 1.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1905 }, { "epoch": 1.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1910 }, { "epoch": 1.36, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1915 }, { "epoch": 1.37, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 319.6012, "step": 1920 }, { "epoch": 1.37, "eval_loss": NaN, "eval_runtime": 15.9336, "eval_samples_per_second": 704.735, "eval_steps_per_second": 88.115, "step": 1920 }, { "epoch": 1.37, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1925 }, { "epoch": 1.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1930 }, { "epoch": 1.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1935 }, { "epoch": 1.38, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1940 }, { "epoch": 1.38, "eval_loss": NaN, "eval_runtime": 16.1212, "eval_samples_per_second": 696.536, "eval_steps_per_second": 87.09, "step": 1940 }, { "epoch": 1.39, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1945 }, { "epoch": 1.39, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1950 }, { "epoch": 1.39, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1955 }, { "epoch": 1.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1960 }, { "epoch": 1.4, "eval_loss": NaN, "eval_runtime": 15.9258, "eval_samples_per_second": 705.083, "eval_steps_per_second": 88.159, "step": 1960 }, { "epoch": 1.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1965 }, { "epoch": 1.4, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1970 }, { "epoch": 1.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1975 }, { "epoch": 1.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1980 }, { "epoch": 1.41, "eval_loss": NaN, "eval_runtime": 15.9, "eval_samples_per_second": 706.226, "eval_steps_per_second": 88.302, "step": 1980 }, { "epoch": 1.41, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1985 }, { "epoch": 1.42, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1990 }, { "epoch": 1.42, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 1995 }, { "epoch": 1.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2000 }, { "epoch": 1.43, "eval_loss": NaN, "eval_runtime": 16.0929, "eval_samples_per_second": 697.761, "eval_steps_per_second": 87.243, "step": 2000 }, { "epoch": 1.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2005 }, { "epoch": 1.43, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2010 }, { "epoch": 1.44, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2015 }, { "epoch": 1.44, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2020 }, { "epoch": 1.44, "eval_loss": NaN, "eval_runtime": 15.9447, "eval_samples_per_second": 704.245, "eval_steps_per_second": 88.054, "step": 2020 }, { "epoch": 1.44, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2025 }, { "epoch": 1.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2030 }, { "epoch": 1.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2035 }, { "epoch": 1.45, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2040 }, { "epoch": 1.45, "eval_loss": NaN, "eval_runtime": 16.1422, "eval_samples_per_second": 695.629, "eval_steps_per_second": 86.977, "step": 2040 }, { "epoch": 1.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2045 }, { "epoch": 1.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2050 }, { "epoch": 1.46, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2055 }, { "epoch": 1.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2060 }, { "epoch": 1.47, "eval_loss": NaN, "eval_runtime": 16.1097, "eval_samples_per_second": 697.032, "eval_steps_per_second": 87.152, "step": 2060 }, { "epoch": 1.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2065 }, { "epoch": 1.47, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2070 }, { "epoch": 1.48, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2075 }, { "epoch": 1.48, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2080 }, { "epoch": 1.48, "eval_loss": NaN, "eval_runtime": 15.9308, "eval_samples_per_second": 704.861, "eval_steps_per_second": 88.131, "step": 2080 }, { "epoch": 1.49, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2085 }, { "epoch": 1.49, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2090 }, { "epoch": 1.49, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2095 }, { "epoch": 1.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2100 }, { "epoch": 1.5, "eval_loss": NaN, "eval_runtime": 16.1276, "eval_samples_per_second": 696.258, "eval_steps_per_second": 87.055, "step": 2100 }, { "epoch": 1.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2105 }, { "epoch": 1.5, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2110 }, { "epoch": 1.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2115 }, { "epoch": 1.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2120 }, { "epoch": 1.51, "eval_loss": NaN, "eval_runtime": 16.1367, "eval_samples_per_second": 695.867, "eval_steps_per_second": 87.007, "step": 2120 }, { "epoch": 1.51, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2125 }, { "epoch": 1.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2130 }, { "epoch": 1.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2135 }, { "epoch": 1.52, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2140 }, { "epoch": 1.52, "eval_loss": NaN, "eval_runtime": 15.9441, "eval_samples_per_second": 704.271, "eval_steps_per_second": 88.057, "step": 2140 }, { "epoch": 1.53, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2145 }, { "epoch": 1.53, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2150 }, { "epoch": 1.54, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2155 }, { "epoch": 1.54, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2160 }, { "epoch": 1.54, "eval_loss": NaN, "eval_runtime": 16.1251, "eval_samples_per_second": 696.367, "eval_steps_per_second": 87.069, "step": 2160 }, { "epoch": 1.54, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2165 }, { "epoch": 1.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2170 }, { "epoch": 1.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2175 }, { "epoch": 1.55, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2180 }, { "epoch": 1.55, "eval_loss": NaN, "eval_runtime": 16.1175, "eval_samples_per_second": 696.696, "eval_steps_per_second": 87.11, "step": 2180 }, { "epoch": 1.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2185 }, { "epoch": 1.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2190 }, { "epoch": 1.56, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2195 }, { "epoch": 1.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2200 }, { "epoch": 1.57, "eval_loss": NaN, "eval_runtime": 15.9153, "eval_samples_per_second": 705.549, "eval_steps_per_second": 88.217, "step": 2200 }, { "epoch": 1.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2205 }, { "epoch": 1.57, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2210 }, { "epoch": 1.58, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2215 }, { "epoch": 1.58, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2220 }, { "epoch": 1.58, "eval_loss": NaN, "eval_runtime": 16.1032, "eval_samples_per_second": 697.314, "eval_steps_per_second": 87.188, "step": 2220 }, { "epoch": 1.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2225 }, { "epoch": 1.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2230 }, { "epoch": 1.59, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2235 }, { "epoch": 1.6, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2240 }, { "epoch": 1.6, "eval_loss": NaN, "eval_runtime": 15.9156, "eval_samples_per_second": 705.534, "eval_steps_per_second": 88.215, "step": 2240 }, { "epoch": 1.6, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2245 }, { "epoch": 1.6, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2250 }, { "epoch": 1.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2255 }, { "epoch": 1.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2260 }, { "epoch": 1.61, "eval_loss": NaN, "eval_runtime": 15.9302, "eval_samples_per_second": 704.887, "eval_steps_per_second": 88.134, "step": 2260 }, { "epoch": 1.61, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2265 }, { "epoch": 1.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2270 }, { "epoch": 1.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2275 }, { "epoch": 1.62, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2280 }, { "epoch": 1.62, "eval_loss": NaN, "eval_runtime": 16.1253, "eval_samples_per_second": 696.358, "eval_steps_per_second": 87.068, "step": 2280 }, { "epoch": 1.63, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2285 }, { "epoch": 1.63, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2290 }, { "epoch": 1.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2295 }, { "epoch": 1.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2300 }, { "epoch": 1.64, "eval_loss": NaN, "eval_runtime": 15.9591, "eval_samples_per_second": 703.611, "eval_steps_per_second": 87.975, "step": 2300 }, { "epoch": 1.64, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2305 }, { "epoch": 1.65, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2310 }, { "epoch": 1.65, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2315 }, { "epoch": 1.65, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2320 }, { "epoch": 1.65, "eval_loss": NaN, "eval_runtime": 15.9298, "eval_samples_per_second": 704.905, "eval_steps_per_second": 88.137, "step": 2320 }, { "epoch": 1.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2325 }, { "epoch": 1.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2330 }, { "epoch": 1.66, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2335 }, { "epoch": 1.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2340 }, { "epoch": 1.67, "eval_loss": NaN, "eval_runtime": 16.1356, "eval_samples_per_second": 695.916, "eval_steps_per_second": 87.013, "step": 2340 }, { "epoch": 1.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2345 }, { "epoch": 1.67, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2350 }, { "epoch": 1.68, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2355 }, { "epoch": 1.68, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2360 }, { "epoch": 1.68, "eval_loss": NaN, "eval_runtime": 15.9295, "eval_samples_per_second": 704.92, "eval_steps_per_second": 88.139, "step": 2360 }, { "epoch": 1.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2365 }, { "epoch": 1.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2370 }, { "epoch": 1.69, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2375 }, { "epoch": 1.7, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2380 }, { "epoch": 1.7, "eval_loss": NaN, "eval_runtime": 16.1228, "eval_samples_per_second": 696.466, "eval_steps_per_second": 87.082, "step": 2380 }, { "epoch": 1.7, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2385 }, { "epoch": 1.7, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2390 }, { "epoch": 1.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2395 }, { "epoch": 1.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2400 }, { "epoch": 1.71, "eval_loss": NaN, "eval_runtime": 16.1353, "eval_samples_per_second": 695.926, "eval_steps_per_second": 87.014, "step": 2400 }, { "epoch": 1.71, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2405 }, { "epoch": 1.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2410 }, { "epoch": 1.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2415 }, { "epoch": 1.72, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2420 }, { "epoch": 1.72, "eval_loss": NaN, "eval_runtime": 15.9416, "eval_samples_per_second": 704.385, "eval_steps_per_second": 88.072, "step": 2420 }, { "epoch": 1.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2425 }, { "epoch": 1.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2430 }, { "epoch": 1.73, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2435 }, { "epoch": 1.74, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2440 }, { "epoch": 1.74, "eval_loss": NaN, "eval_runtime": 16.0914, "eval_samples_per_second": 697.828, "eval_steps_per_second": 87.252, "step": 2440 }, { "epoch": 1.74, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2445 }, { "epoch": 1.75, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2450 }, { "epoch": 1.75, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.1965, "step": 2455 }, { "epoch": 1.75, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2460 }, { "epoch": 1.75, "eval_loss": NaN, "eval_runtime": 16.1003, "eval_samples_per_second": 697.441, "eval_steps_per_second": 87.203, "step": 2460 }, { "epoch": 1.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2465 }, { "epoch": 1.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2470 }, { "epoch": 1.76, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2475 }, { "epoch": 1.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2480 }, { "epoch": 1.77, "eval_loss": NaN, "eval_runtime": 15.8994, "eval_samples_per_second": 706.254, "eval_steps_per_second": 88.305, "step": 2480 }, { "epoch": 1.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2485 }, { "epoch": 1.77, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2490 }, { "epoch": 1.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2495 }, { "epoch": 1.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2500 }, { "epoch": 1.78, "eval_loss": NaN, "eval_runtime": 16.1, "eval_samples_per_second": 697.452, "eval_steps_per_second": 87.205, "step": 2500 }, { "epoch": 1.78, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2505 }, { "epoch": 1.79, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2510 }, { "epoch": 1.79, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2515 }, { "epoch": 1.8, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2520 }, { "epoch": 1.8, "eval_loss": NaN, "eval_runtime": 15.8921, "eval_samples_per_second": 706.575, "eval_steps_per_second": 88.346, "step": 2520 }, { "epoch": 1.8, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2525 }, { "epoch": 1.8, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2530 }, { "epoch": 1.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2535 }, { "epoch": 1.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2540 }, { "epoch": 1.81, "eval_loss": NaN, "eval_runtime": 15.9033, "eval_samples_per_second": 706.078, "eval_steps_per_second": 88.283, "step": 2540 }, { "epoch": 1.81, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2545 }, { "epoch": 1.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2550 }, { "epoch": 1.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2555 }, { "epoch": 1.82, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2560 }, { "epoch": 1.82, "eval_loss": NaN, "eval_runtime": 16.0359, "eval_samples_per_second": 700.244, "eval_steps_per_second": 87.554, "step": 2560 }, { "epoch": 1.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2565 }, { "epoch": 1.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2570 }, { "epoch": 1.83, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2575 }, { "epoch": 1.84, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2580 }, { "epoch": 1.84, "eval_loss": NaN, "eval_runtime": 15.8356, "eval_samples_per_second": 709.1, "eval_steps_per_second": 88.661, "step": 2580 }, { "epoch": 1.84, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2585 }, { "epoch": 1.85, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2590 }, { "epoch": 1.85, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2595 }, { "epoch": 1.85, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2600 }, { "epoch": 1.85, "eval_loss": NaN, "eval_runtime": 15.865, "eval_samples_per_second": 707.785, "eval_steps_per_second": 88.497, "step": 2600 }, { "epoch": 1.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2605 }, { "epoch": 1.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2610 }, { "epoch": 1.86, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2615 }, { "epoch": 1.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2620 }, { "epoch": 1.87, "eval_loss": NaN, "eval_runtime": 16.0167, "eval_samples_per_second": 701.083, "eval_steps_per_second": 87.659, "step": 2620 }, { "epoch": 1.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2625 }, { "epoch": 1.87, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2630 }, { "epoch": 1.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2635 }, { "epoch": 1.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2640 }, { "epoch": 1.88, "eval_loss": NaN, "eval_runtime": 15.8289, "eval_samples_per_second": 709.401, "eval_steps_per_second": 88.699, "step": 2640 }, { "epoch": 1.88, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2645 }, { "epoch": 1.89, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2650 }, { "epoch": 1.89, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2655 }, { "epoch": 1.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2660 }, { "epoch": 1.9, "eval_loss": NaN, "eval_runtime": 15.9974, "eval_samples_per_second": 701.925, "eval_steps_per_second": 87.764, "step": 2660 }, { "epoch": 1.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2665 }, { "epoch": 1.9, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2670 }, { "epoch": 1.91, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2675 }, { "epoch": 1.91, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2680 }, { "epoch": 1.91, "eval_loss": NaN, "eval_runtime": 16.0607, "eval_samples_per_second": 699.16, "eval_steps_per_second": 87.418, "step": 2680 }, { "epoch": 1.91, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2685 }, { "epoch": 1.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2690 }, { "epoch": 1.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2695 }, { "epoch": 1.92, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2700 }, { "epoch": 1.92, "eval_loss": NaN, "eval_runtime": 15.8531, "eval_samples_per_second": 708.316, "eval_steps_per_second": 88.563, "step": 2700 }, { "epoch": 1.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2705 }, { "epoch": 1.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2710 }, { "epoch": 1.93, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2715 }, { "epoch": 1.94, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2720 }, { "epoch": 1.94, "eval_loss": NaN, "eval_runtime": 16.0405, "eval_samples_per_second": 700.043, "eval_steps_per_second": 87.529, "step": 2720 }, { "epoch": 1.94, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2725 }, { "epoch": 1.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2730 }, { "epoch": 1.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2735 }, { "epoch": 1.95, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2740 }, { "epoch": 1.95, "eval_loss": NaN, "eval_runtime": 16.0283, "eval_samples_per_second": 700.572, "eval_steps_per_second": 87.595, "step": 2740 }, { "epoch": 1.96, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2745 }, { "epoch": 1.96, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2750 }, { "epoch": 1.96, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2755 }, { "epoch": 1.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2760 }, { "epoch": 1.97, "eval_loss": NaN, "eval_runtime": 15.8417, "eval_samples_per_second": 708.827, "eval_steps_per_second": 88.627, "step": 2760 }, { "epoch": 1.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2765 }, { "epoch": 1.97, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2770 }, { "epoch": 1.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2775 }, { "epoch": 1.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2780 }, { "epoch": 1.98, "eval_loss": NaN, "eval_runtime": 16.0698, "eval_samples_per_second": 698.762, "eval_steps_per_second": 87.369, "step": 2780 }, { "epoch": 1.98, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2785 }, { "epoch": 1.99, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2790 }, { "epoch": 1.99, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2795 }, { "epoch": 2.0, "grad_norm": NaN, "learning_rate": 0.0003, "loss": 0.0, "step": 2800 }, { "epoch": 2.0, "eval_loss": NaN, "eval_runtime": 15.8642, "eval_samples_per_second": 707.819, "eval_steps_per_second": 88.501, "step": 2800 } ], "logging_steps": 5, "max_steps": 2806, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 20, "total_flos": 3171114172809216.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }