| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0120603862697142, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "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": 0.7047, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6771, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.676, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6731, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6757, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6726, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6734, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6718, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6678, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6683, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6735, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6669, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6719, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6684, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6675, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6806, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6743, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6735, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6739, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6739, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6761, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6714, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6707, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6651, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6757, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6759, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6698, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6687, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6697, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6678, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6604, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6669, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6655, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6569, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6666, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6647, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6612, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6616, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6607, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6596, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6542, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6606, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.654, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6631, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6558, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6498, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6728, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6649, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6569, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6582, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6484, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6476, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6505, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6523, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6521, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6467, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6628, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6523, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.655, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6496, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6498, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6498, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6542, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6445, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6511, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6482, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6462, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6623, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6473, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.651, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6453, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6508, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6541, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6526, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6391, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6467, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6505, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6405, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6533, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6432, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6448, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6361, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6416, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6381, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6461, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6439, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6433, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6468, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6507, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6481, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6442, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6384, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6418, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6505, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.637, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6381, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6377, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6737, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.64, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.641, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6335, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6583, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6388, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6357, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6326, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6349, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6468, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6384, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6427, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6454, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6379, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6352, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6299, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6381, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6322, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.653, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6367, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6477, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6413, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6374, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6359, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6325, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6454, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 5e-05, | |
| "loss": 0.641, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6332, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6338, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6351, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6365, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6376, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6367, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6386, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6433, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6377, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6389, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6298, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6304, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6361, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6301, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6303, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6299, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6376, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6315, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6447, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6349, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6343, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6325, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6343, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6389, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6359, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6302, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6329, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6315, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6237, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6347, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6293, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6428, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6466, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6391, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6371, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6357, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6323, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6239, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6316, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.638, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6391, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6494, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6448, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6454, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6521, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6653, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6432, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6284, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6347, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6373, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.635, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6322, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6297, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6532, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.644, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6428, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6344, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6268, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6273, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6372, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6398, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6491, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6318, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6339, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6355, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6381, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6242, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6314, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6327, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6354, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.627, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6383, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.63, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6313, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6367, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6393, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.644, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.644, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6387, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6388, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6337, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6284, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6292, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6267, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6312, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.634, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6267, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6227, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6319, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.627, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.627, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6425, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6301, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6283, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6311, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6316, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6277, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6399, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6324, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6268, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6306, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6258, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6203, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6305, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6385, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6288, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6283, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6349, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.625, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6223, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6382, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6345, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6462, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6367, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6358, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6357, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6338, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6255, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6199, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6252, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6258, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6235, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6328, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6359, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6329, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.624, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6463, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6195, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6446, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6264, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6234, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6266, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6209, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6213, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6226, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.627, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6224, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6212, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6141, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6249, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6204, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6259, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6206, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6241, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6237, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.621, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.622, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6217, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6206, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6366, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6341, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6414, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6208, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6218, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6278, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6295, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6194, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6278, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6248, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6213, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6219, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6362, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6377, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6333, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6526, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6221, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6274, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6226, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6217, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6222, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6246, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6311, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6182, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6112, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6145, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6233, | |
| "step": 3000 | |
| } | |
| ], | |
| "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 | |
| } | |