| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.6596306068601583, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 4.989006156552331e-05, | |
| "loss": 1.8193, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 4.978012313104662e-05, | |
| "loss": 1.7656, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 4.967018469656992e-05, | |
| "loss": 1.7573, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.956024626209323e-05, | |
| "loss": 1.6758, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.9450307827616535e-05, | |
| "loss": 1.5776, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 4.934036939313985e-05, | |
| "loss": 1.615, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.9230430958663146e-05, | |
| "loss": 1.5378, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.912049252418646e-05, | |
| "loss": 1.6188, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 4.901055408970976e-05, | |
| "loss": 1.5489, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 4.8900615655233075e-05, | |
| "loss": 1.5724, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 4.879067722075638e-05, | |
| "loss": 1.5262, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.8680738786279686e-05, | |
| "loss": 1.5224, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.857080035180299e-05, | |
| "loss": 1.5176, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.84608619173263e-05, | |
| "loss": 1.5263, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.835092348284961e-05, | |
| "loss": 1.5832, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.824098504837291e-05, | |
| "loss": 1.5403, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.813104661389622e-05, | |
| "loss": 1.5689, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.802110817941953e-05, | |
| "loss": 1.5389, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.7911169744942836e-05, | |
| "loss": 1.5593, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.780123131046614e-05, | |
| "loss": 1.52, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.7691292875989446e-05, | |
| "loss": 1.6009, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.758135444151275e-05, | |
| "loss": 1.5612, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.7471416007036063e-05, | |
| "loss": 1.5532, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.736147757255937e-05, | |
| "loss": 1.5636, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.7251539138082674e-05, | |
| "loss": 1.56, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.714160070360598e-05, | |
| "loss": 1.5406, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.703166226912929e-05, | |
| "loss": 1.4851, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.6921723834652596e-05, | |
| "loss": 1.5333, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.681178540017591e-05, | |
| "loss": 1.5684, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.670184696569921e-05, | |
| "loss": 1.5636, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.659190853122252e-05, | |
| "loss": 1.5434, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.6481970096745824e-05, | |
| "loss": 1.5002, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.6372031662269136e-05, | |
| "loss": 1.5413, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.6262093227792435e-05, | |
| "loss": 1.5275, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.615215479331575e-05, | |
| "loss": 1.508, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.604221635883905e-05, | |
| "loss": 1.5033, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.593227792436236e-05, | |
| "loss": 1.5416, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.582233948988567e-05, | |
| "loss": 1.5092, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.5712401055408974e-05, | |
| "loss": 1.5605, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.560246262093228e-05, | |
| "loss": 1.5624, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5492524186455585e-05, | |
| "loss": 1.5394, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.53825857519789e-05, | |
| "loss": 1.476, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.52726473175022e-05, | |
| "loss": 1.5367, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.516270888302551e-05, | |
| "loss": 1.5041, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.505277044854881e-05, | |
| "loss": 1.5342, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.4942832014072125e-05, | |
| "loss": 1.5068, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.483289357959543e-05, | |
| "loss": 1.4912, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.4722955145118735e-05, | |
| "loss": 1.4893, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.461301671064204e-05, | |
| "loss": 1.5338, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.450307827616535e-05, | |
| "loss": 1.5258, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.439313984168866e-05, | |
| "loss": 1.5373, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.428320140721196e-05, | |
| "loss": 1.4972, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.417326297273527e-05, | |
| "loss": 1.4868, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.406332453825858e-05, | |
| "loss": 1.5167, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.3953386103781885e-05, | |
| "loss": 1.5242, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 4.384344766930519e-05, | |
| "loss": 1.5309, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4.3733509234828496e-05, | |
| "loss": 1.5091, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4.36235708003518e-05, | |
| "loss": 1.5041, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 4.351363236587511e-05, | |
| "loss": 1.4934, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.340369393139842e-05, | |
| "loss": 1.5113, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.329375549692173e-05, | |
| "loss": 1.4899, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.318381706244503e-05, | |
| "loss": 1.5019, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 4.307387862796834e-05, | |
| "loss": 1.5019, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 4.2963940193491646e-05, | |
| "loss": 1.5331, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.285400175901496e-05, | |
| "loss": 1.5444, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 4.2744063324538256e-05, | |
| "loss": 1.527, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 4.263412489006157e-05, | |
| "loss": 1.5329, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 4.2524186455584874e-05, | |
| "loss": 1.5572, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.2414248021108186e-05, | |
| "loss": 1.53, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.230430958663149e-05, | |
| "loss": 1.4891, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.2194371152154796e-05, | |
| "loss": 1.5004, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.20844327176781e-05, | |
| "loss": 1.5134, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 4.1974494283201407e-05, | |
| "loss": 1.5081, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 4.186455584872472e-05, | |
| "loss": 1.5006, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 4.1754617414248024e-05, | |
| "loss": 1.515, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 4.164467897977133e-05, | |
| "loss": 1.4862, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 4.1534740545294634e-05, | |
| "loss": 1.5196, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 4.1424802110817946e-05, | |
| "loss": 1.4908, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 4.131486367634125e-05, | |
| "loss": 1.5443, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.120492524186456e-05, | |
| "loss": 1.5457, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.109498680738786e-05, | |
| "loss": 1.5616, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 4.0985048372911174e-05, | |
| "loss": 1.5537, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.087510993843448e-05, | |
| "loss": 1.5206, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.0765171503957784e-05, | |
| "loss": 1.4923, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 4.065523306948109e-05, | |
| "loss": 1.4721, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.05452946350044e-05, | |
| "loss": 1.5357, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.043535620052771e-05, | |
| "loss": 1.5163, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.032541776605102e-05, | |
| "loss": 1.5406, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 4.021547933157432e-05, | |
| "loss": 1.5151, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 4.010554089709763e-05, | |
| "loss": 1.5482, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 3.9995602462620935e-05, | |
| "loss": 1.487, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 3.988566402814424e-05, | |
| "loss": 1.5559, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 3.9775725593667545e-05, | |
| "loss": 1.5085, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 3.966578715919085e-05, | |
| "loss": 1.5091, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 3.955584872471416e-05, | |
| "loss": 1.512, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 3.944591029023747e-05, | |
| "loss": 1.4824, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 3.933597185576078e-05, | |
| "loss": 1.4935, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 3.922603342128408e-05, | |
| "loss": 1.5034, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 3.911609498680739e-05, | |
| "loss": 1.5449, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 3.9006156552330695e-05, | |
| "loss": 1.481, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4548, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 6.3202482192384e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |