| { | |
| "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": 0.7081, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6774, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6781, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6772, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6802, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6749, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6745, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6729, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5e-05, | |
| "loss": 0.636, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6237, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6098, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5978, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.602, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5959, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5906, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5778, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5795, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5883, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.604, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5882, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5747, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5745, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5846, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5823, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5722, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5599, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5716, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5832, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5745, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5691, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5711, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5821, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5557, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5581, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5761, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5643, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5745, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5702, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5648, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5691, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5e-05, | |
| "loss": 0.556, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5578, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.564, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5562, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.559, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.565, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5508, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5588, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5584, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5555, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5597, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5483, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5621, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5568, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5527, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5524, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5458, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5426, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5524, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5553, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5582, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5518, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5485, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5511, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5444, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5522, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5417, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5543, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.56, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5394, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5475, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5437, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5438, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5454, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5481, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5434, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5422, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5465, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5447, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5396, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.543, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5461, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5482, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5426, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5453, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5493, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5477, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.542, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5393, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5365, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5433, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 5e-05, | |
| "loss": 0.542, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5368, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.542, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5467, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5382, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5366, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5354, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5322, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5379, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5398, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5337, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5384, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5256, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5353, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5233, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5483, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5362, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5426, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5347, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5377, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5374, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5e-05, | |
| "loss": 0.532, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5315, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5274, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.537, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5427, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5288, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5259, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.531, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5263, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5305, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5245, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5306, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.526, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5308, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5226, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5295, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5297, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5293, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5387, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5252, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5e-05, | |
| "loss": 0.524, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5231, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5248, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5332, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5366, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5231, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5206, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5237, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5201, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5366, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5336, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5282, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5269, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5245, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5203, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5225, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5311, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5229, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5266, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5283, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5279, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5299, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5267, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5186, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5238, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5249, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5225, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5226, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5188, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.525, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5249, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5225, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5281, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5204, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5193, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5278, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5224, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5218, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5171, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5239, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5231, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5164, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5181, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5166, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5179, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5169, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5275, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5179, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5216, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5236, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5125, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-05, | |
| "loss": 0.527, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5224, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5304, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5242, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5351, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5262, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5178, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5104, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5172, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 5e-05, | |
| "loss": 0.517, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5154, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5285, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5163, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5104, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5228, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5131, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5173, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5137, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5132, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5154, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5194, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 5e-05, | |
| "loss": 0.521, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5166, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5161, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5143, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5133, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5188, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5099, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5181, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5148, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5155, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5108, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5026, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5187, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5156, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5164, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5129, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5143, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5206, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.508, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5101, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5069, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5082, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5206, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5134, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5099, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5114, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5132, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5125, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5055, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5e-05, | |
| "loss": 0.514, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5093, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.506, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5198, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5187, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5112, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5189, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5104, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5105, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5165, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5088, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.504, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5065, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5053, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5135, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5078, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.508, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5038, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5114, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5055, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5011, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5025, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5113, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.506, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5031, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.505, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4966, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5132, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4975, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5108, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.508, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.514, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5071, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4952, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5021, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5112, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5037, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5106, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5011, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5103, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5004, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5045, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5019, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4928, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5029, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5175, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5114, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5021, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5084, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5008, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5102, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5133, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.508, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5054, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5191, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5051, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4943, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5044, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4943, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5e-05, | |
| "loss": 0.504, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4978, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.492, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4938, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5005, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4997, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4968, | |
| "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 | |
| } | |