| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3373534620899047, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5854, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5466, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5426, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5393, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.55, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5355, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5387, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5299, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5183, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5188, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5186, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5135, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5221, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5163, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5063, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5134, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5058, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5059, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5077, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5089, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5075, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5017, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5171, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5022, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5005, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5025, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5046, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5174, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5037, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4952, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4914, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4867, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4998, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.482, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.498, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4982, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4931, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4938, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4859, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4852, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4942, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5014, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5076, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4962, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4802, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5002, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4979, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4999, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4908, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4884, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4801, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.478, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4838, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4879, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4761, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4853, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4938, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.477, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.491, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4678, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4814, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4822, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4792, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4791, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4839, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4846, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4833, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4821, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4778, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4783, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4644, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4771, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4747, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4802, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4738, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4768, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4692, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4684, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4775, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4714, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4744, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4752, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4727, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4693, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4746, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4699, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4694, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4745, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4674, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4754, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4796, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4708, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4746, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4768, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4709, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.466, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4706, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4756, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4687, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4667, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 8892, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 1000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |