[oftrain] device=Cuda(0) [train] spawning 256 env workers across 8 shard(s) [Cuda(0), Cuda(1), Cuda(2), Cuda(3), Cuda(4), Cuda(5), Cuda(6), Cuda(7)] (stage=0, max_ticks=3000)... [train] all 256 envs ready [train] policy params: 11166694 per shard x 8 shard(s) on [Cuda(0), Cuda(1), Cuda(2), Cuda(3), Cuda(4), Cuda(5), Cuda(6), Cuda(7)] [update 0] steps/s= 838.9 decisions_total= 16384 eps_done= 1 recent_reward= 3.496 pg=-0.0136 v=0.2101 ent=2.889 entq=-0.103 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=29s update_s=19.5 collect_s=19.5 train_s=19.0 gpu_mem%=87 min_mean_util%=37 [gpu0=25% gpu1=2% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=0%] [update 1] steps/s= 905.5 decisions_total= 32768 eps_done= 6 recent_reward= 3.341 pg=+0.0160 v=0.0870 ent=3.649 entq=-0.079 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=48s update_s=18.1 collect_s=18.1 train_s=16.5 gpu_mem%=88 min_mean_util%=44 [gpu0=0% gpu1=15% gpu2=0% gpu3=2% gpu4=0% gpu5=30% gpu6=0% gpu7=25%] [update 2] steps/s= 945.1 decisions_total= 49152 eps_done= 256 recent_reward= 6.405 pg=-0.1168 v=0.1797 ent=4.284 entq=-0.103 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=65s update_s=17.3 collect_s=17.3 train_s=15.3 gpu_mem%=88 min_mean_util%=47 [gpu0=25% gpu1=29% gpu2=9% gpu3=11% gpu4=0% gpu5=0% gpu6=13% gpu7=25%] [update 3] steps/s= 902.6 decisions_total= 65536 eps_done= 256 recent_reward= 6.405 pg=+0.3740 v=2.3871 ent=3.344 entq=-0.096 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=83s update_s=18.2 collect_s=18.0 train_s=16.4 gpu_mem%=88 min_mean_util%=49 [gpu0=0% gpu1=0% gpu2=25% gpu3=25% gpu4=0% gpu5=0% gpu6=0% gpu7=14%] [update 4] steps/s= 885.7 decisions_total= 81920 eps_done= 262 recent_reward= 6.438 pg=+0.0549 v=0.4944 ent=4.224 entq=-0.091 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=102s update_s=18.5 collect_s=18.3 train_s=17.2 gpu_mem%=88 min_mean_util%=51 [gpu0=0% gpu1=24% gpu2=29% gpu3=25% gpu4=0% gpu5=0% gpu6=25% gpu7=26%] [update 5] steps/s= 856.4 decisions_total= 98304 eps_done= 512 recent_reward= 7.386 pg=+0.3175 v=731.0077 ent=1.778 entq=-1.392 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=121s update_s=19.1 collect_s=19.0 train_s=18.8 gpu_mem%=88 min_mean_util%=53 [gpu0=45% gpu1=99% gpu2=77% gpu3=53% gpu4=77% gpu5=93% gpu6=100% gpu7=100%] [update 6] steps/s= 837.8 decisions_total= 114688 eps_done= 512 recent_reward= 7.386 pg=+0.4172 v=7.6023 ent=2.310 entq=-0.084 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=140s update_s=19.6 collect_s=19.5 train_s=18.4 gpu_mem%=88 min_mean_util%=53 [gpu0=25% gpu1=0% gpu2=0% gpu3=0% gpu4=11% gpu5=0% gpu6=26% gpu7=24%] [update 7] steps/s= 864.6 decisions_total= 131072 eps_done= 518 recent_reward= 7.404 pg=+0.0582 v=42.4978 ent=0.410 entq=-0.483 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=159s update_s=19.0 collect_s=18.9 train_s=16.9 gpu_mem%=88 min_mean_util%=54 [gpu0=24% gpu1=28% gpu2=0% gpu3=10% gpu4=25% gpu5=27% gpu6=23% gpu7=26%] [update 8] steps/s= 866.0 decisions_total= 147456 eps_done= 768 recent_reward= 6.769 pg=+0.0968 v=5.4085 ent=0.961 entq=-0.443 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=178s update_s=18.9 collect_s=18.9 train_s=18.1 gpu_mem%=88 min_mean_util%=54 [gpu0=28% gpu1=21% gpu2=28% gpu3=0% gpu4=23% gpu5=25% gpu6=27% gpu7=2%] [update 9] steps/s= 896.1 decisions_total= 163840 eps_done= 769 recent_reward= 6.695 pg=+0.3513 v=66.1757 ent=0.514 entq=-0.132 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=196s update_s=18.3 collect_s=18.2 train_s=15.8 gpu_mem%=88 min_mean_util%=54 [gpu0=23% gpu1=26% gpu2=0% gpu3=28% gpu4=28% gpu5=0% gpu6=0% gpu7=31%] [update 10] steps/s= 843.6 decisions_total= 180224 eps_done= 776 recent_reward= 6.518 pg=+0.5868 v=15.0915 ent=1.409 entq=-0.915 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=216s update_s=19.4 collect_s=19.4 train_s=16.4 gpu_mem%=88 min_mean_util%=54 [gpu0=28% gpu1=13% gpu2=0% gpu3=0% gpu4=0% gpu5=26% gpu6=0% gpu7=21%] [update 11] steps/s= 851.6 decisions_total= 196608 eps_done= 1024 recent_reward= 8.055 pg=+0.1619 v=42.1973 ent=0.331 entq=-1.118 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=235s update_s=19.2 collect_s=19.2 train_s=16.9 gpu_mem%=88 min_mean_util%=54 [gpu0=26% gpu1=28% gpu2=0% gpu3=0% gpu4=0% gpu5=28% gpu6=0% gpu7=0%] [update 12] steps/s= 844.3 decisions_total= 212992 eps_done= 1025 recent_reward= 8.028 pg=+0.3512 v=14.0555 ent=1.862 entq=-0.203 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=254s update_s=19.4 collect_s=19.4 train_s=17.3 gpu_mem%=88 min_mean_util%=53 [gpu0=28% gpu1=0% gpu2=26% gpu3=30% gpu4=22% gpu5=0% gpu6=28% gpu7=0%] [update 13] steps/s= 867.8 decisions_total= 229376 eps_done= 1032 recent_reward= 7.631 pg=+0.5421 v=15.2264 ent=2.821 entq=-0.607 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=273s update_s=18.9 collect_s=18.8 train_s=17.0 gpu_mem%=89 min_mean_util%=54 [gpu0=26% gpu1=28% gpu2=16% gpu3=25% gpu4=26% gpu5=24% gpu6=28% gpu7=11%] [update 14] steps/s= 892.2 decisions_total= 245760 eps_done= 1280 recent_reward= 5.488 pg=+0.2646 v=212.4965 ent=2.227 entq=-0.573 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=292s update_s=18.4 collect_s=18.3 train_s=16.1 gpu_mem%=89 min_mean_util%=54 [gpu0=33% gpu1=27% gpu2=26% gpu3=20% gpu4=29% gpu5=25% gpu6=30% gpu7=25%] [update 15] steps/s= 904.8 decisions_total= 262144 eps_done= 1282 recent_reward= 5.430 pg=+0.7461 v=151.3084 ent=1.762 entq=-0.990 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=310s update_s=18.1 collect_s=18.1 train_s=17.2 gpu_mem%=91 min_mean_util%=54 [gpu0=15% gpu1=25% gpu2=28% gpu3=29% gpu4=30% gpu5=31% gpu6=28% gpu7=15%] [update 16] steps/s= 861.5 decisions_total= 278528 eps_done= 1297 recent_reward= 5.244 pg=+0.5019 v=3.2465 ent=2.868 entq=-0.177 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=329s update_s=19.0 collect_s=19.0 train_s=16.7 gpu_mem%=91 min_mean_util%=54 [gpu0=28% gpu1=0% gpu2=0% gpu3=36% gpu4=7% gpu5=12% gpu6=12% gpu7=28%] [update 17] steps/s= 903.9 decisions_total= 294912 eps_done= 1536 recent_reward= 5.785 pg=+0.1446 v=51.4666 ent=3.921 entq=-0.336 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=347s update_s=18.1 collect_s=18.1 train_s=17.0 gpu_mem%=91 min_mean_util%=54 [gpu0=12% gpu1=26% gpu2=25% gpu3=19% gpu4=21% gpu5=14% gpu6=25% gpu7=15%] [update 18] steps/s= 870.3 decisions_total= 311296 eps_done= 1537 recent_reward= 5.784 pg=+0.4625 v=1393.3689 ent=2.558 entq=-0.199 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=366s update_s=18.8 collect_s=18.8 train_s=18.4 gpu_mem%=91 min_mean_util%=54 [gpu0=0% gpu1=33% gpu2=24% gpu3=2% gpu4=2% gpu5=11% gpu6=2% gpu7=17%] [update 19] steps/s= 888.9 decisions_total= 327680 eps_done= 1562 recent_reward= 5.583 pg=+0.1290 v=86.0014 ent=1.946 entq=-0.973 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=384s update_s=18.4 collect_s=18.4 train_s=17.4 gpu_mem%=91 min_mean_util%=54 [gpu0=11% gpu1=0% gpu2=12% gpu3=0% gpu4=11% gpu5=0% gpu6=0% gpu7=30%] [update 20] steps/s= 892.7 decisions_total= 344064 eps_done= 1792 recent_reward= 6.057 pg=+0.2472 v=77.0128 ent=5.619 entq=-0.059 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=403s update_s=18.4 collect_s=18.3 train_s=18.0 gpu_mem%=91 min_mean_util%=54 [gpu0=25% gpu1=0% gpu2=2% gpu3=27% gpu4=28% gpu5=2% gpu6=25% gpu7=14%] [update 21] steps/s= 916.9 decisions_total= 360448 eps_done= 1795 recent_reward= 5.944 pg=+0.2665 v=1298.1292 ent=3.764 entq=-1.944 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=420s update_s=17.9 collect_s=17.8 train_s=16.4 gpu_mem%=91 min_mean_util%=54 [gpu0=37% gpu1=23% gpu2=30% gpu3=26% gpu4=0% gpu5=0% gpu6=32% gpu7=0%] [update 22] steps/s= 895.8 decisions_total= 376832 eps_done= 1817 recent_reward= 5.620 pg=+0.6737 v=22.0915 ent=4.608 entq=-0.089 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=439s update_s=18.3 collect_s=18.2 train_s=16.4 gpu_mem%=91 min_mean_util%=54 [gpu0=0% gpu1=4% gpu2=0% gpu3=5% gpu4=23% gpu5=30% gpu6=0% gpu7=0%] [update 23] steps/s= 914.0 decisions_total= 393216 eps_done= 2048 recent_reward= 5.761 pg=+0.1554 v=617.6539 ent=0.973 entq=-1.660 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=457s update_s=17.9 collect_s=17.9 train_s=15.8 gpu_mem%=91 min_mean_util%=55 [gpu0=22% gpu1=27% gpu2=27% gpu3=21% gpu4=19% gpu5=25% gpu6=0% gpu7=26%] [update 24] steps/s= 851.0 decisions_total= 409600 eps_done= 2050 recent_reward= 5.782 pg=+0.3644 v=482.2890 ent=1.663 entq=-0.136 ecoef=0.0100 stage=0 lr=3.00e-4 elapsed=476s update_s=19.3 collect_s=19.2 train_s=16.8 gpu_mem%=91 min_mean_util%=54 [gpu0=21% gpu1=29% gpu2=12% gpu3=0% gpu4=0% gpu5=0% gpu6=28% gpu7=0%] [update 25] steps/s= 886.6 decisions_total= 425984 eps_done= 2073 recent_reward= 6.230 pg=+0.1420 v=701.1448 ent=1.116 entq=-0.014 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=494s update_s=18.5 collect_s=18.4 train_s=16.1 gpu_mem%=92 min_mean_util%=54 [gpu0=15% gpu1=35% gpu2=28% gpu3=35% gpu4=25% gpu5=29% gpu6=34% gpu7=28%] [train] checkpoint saved: checkpoints/policy_update25.ot [update 26] steps/s= 875.3 decisions_total= 442368 eps_done= 2118 recent_reward= 6.089 pg=+0.0424 v=18.3161 ent=1.431 entq=-0.009 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=513s update_s=18.7 collect_s=18.7 train_s=16.3 gpu_mem%=91 min_mean_util%=55 [gpu0=24% gpu1=19% gpu2=29% gpu3=0% gpu4=29% gpu5=10% gpu6=27% gpu7=0%] [update 27] steps/s= 869.3 decisions_total= 458752 eps_done= 2307 recent_reward= 6.378 pg=+0.1266 v=36.6137 ent=0.364 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=532s update_s=18.8 collect_s=18.8 train_s=16.9 gpu_mem%=91 min_mean_util%=55 [gpu0=27% gpu1=38% gpu2=35% gpu3=0% gpu4=26% gpu5=24% gpu6=0% gpu7=0%] [update 28] steps/s= 843.0 decisions_total= 475136 eps_done= 2327 recent_reward= 6.748 pg=+0.1959 v=68.7315 ent=0.527 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=551s update_s=19.4 collect_s=19.4 train_s=17.2 gpu_mem%=91 min_mean_util%=55 [gpu0=29% gpu1=0% gpu2=0% gpu3=0% gpu4=27% gpu5=0% gpu6=25% gpu7=0%] [update 29] steps/s= 848.5 decisions_total= 491520 eps_done= 2369 recent_reward= 7.453 pg=+0.0537 v=25.9977 ent=1.140 entq=-0.052 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=571s update_s=19.3 collect_s=19.3 train_s=16.9 gpu_mem%=91 min_mean_util%=55 [gpu0=33% gpu1=27% gpu2=26% gpu3=29% gpu4=26% gpu5=27% gpu6=27% gpu7=21%] [update 30] steps/s= 868.7 decisions_total= 507904 eps_done= 2562 recent_reward= 7.848 pg=+0.0760 v=23.2998 ent=1.295 entq=-0.097 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=590s update_s=18.9 collect_s=18.8 train_s=16.6 gpu_mem%=91 min_mean_util%=55 [gpu0=29% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=33% gpu6=8% gpu7=0%] [update 31] steps/s= 850.1 decisions_total= 524288 eps_done= 2580 recent_reward= 7.640 pg=+0.2288 v=137.1612 ent=1.083 entq=-0.005 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=609s update_s=19.3 collect_s=19.2 train_s=17.2 gpu_mem%=91 min_mean_util%=55 [gpu0=12% gpu1=25% gpu2=0% gpu3=0% gpu4=32% gpu5=23% gpu6=24% gpu7=0%] [update 32] steps/s= 847.5 decisions_total= 540672 eps_done= 2620 recent_reward= 7.013 pg=+0.1280 v=70.6863 ent=1.296 entq=-0.085 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=628s update_s=19.3 collect_s=19.3 train_s=18.4 gpu_mem%=91 min_mean_util%=55 [gpu0=25% gpu1=25% gpu2=21% gpu3=26% gpu4=37% gpu5=0% gpu6=0% gpu7=12%] [update 33] steps/s= 859.8 decisions_total= 557056 eps_done= 2818 recent_reward= 10.335 pg=+0.1906 v=360.5235 ent=0.746 entq=-0.023 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=647s update_s=19.1 collect_s=19.0 train_s=17.4 gpu_mem%=91 min_mean_util%=55 [gpu0=0% gpu1=0% gpu2=0% gpu3=0% gpu4=36% gpu5=0% gpu6=0% gpu7=27%] [update 34] steps/s= 851.6 decisions_total= 573440 eps_done= 2832 recent_reward= 9.129 pg=+0.1788 v=47.4989 ent=1.153 entq=-0.047 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=666s update_s=19.2 collect_s=19.2 train_s=18.5 gpu_mem%=91 min_mean_util%=55 [gpu0=30% gpu1=33% gpu2=37% gpu3=25% gpu4=26% gpu5=25% gpu6=0% gpu7=0%] [update 35] steps/s= 864.8 decisions_total= 589824 eps_done= 2871 recent_reward= 6.311 pg=+0.0749 v=158.5300 ent=1.710 entq=-0.005 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=685s update_s=18.9 collect_s=18.9 train_s=17.3 gpu_mem%=91 min_mean_util%=55 [gpu0=26% gpu1=28% gpu2=29% gpu3=26% gpu4=22% gpu5=11% gpu6=31% gpu7=12%] [update 36] steps/s= 838.1 decisions_total= 606208 eps_done= 3074 recent_reward= 6.022 pg=+0.0392 v=25.4236 ent=2.246 entq=-0.031 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=705s update_s=19.5 collect_s=19.5 train_s=16.8 gpu_mem%=91 min_mean_util%=55 [gpu0=32% gpu1=0% gpu2=30% gpu3=29% gpu4=26% gpu5=15% gpu6=23% gpu7=0%] [update 37] steps/s= 845.7 decisions_total= 622592 eps_done= 3087 recent_reward= 6.665 pg=+0.0303 v=15.1308 ent=0.905 entq=-0.016 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=724s update_s=19.4 collect_s=19.3 train_s=18.0 gpu_mem%=91 min_mean_util%=55 [gpu0=26% gpu1=12% gpu2=11% gpu3=29% gpu4=24% gpu5=35% gpu6=0% gpu7=0%] [update 38] steps/s= 865.2 decisions_total= 638976 eps_done= 3122 recent_reward= 7.927 pg=+0.5757 v=30.0962 ent=2.378 entq=-0.328 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=743s update_s=18.9 collect_s=18.9 train_s=16.4 gpu_mem%=91 min_mean_util%=55 [gpu0=24% gpu1=29% gpu2=28% gpu3=31% gpu4=0% gpu5=26% gpu6=0% gpu7=0%] [update 39] steps/s= 863.2 decisions_total= 655360 eps_done= 3328 recent_reward= 9.606 pg=+0.0776 v=26.5073 ent=1.947 entq=-0.001 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=762s update_s=19.0 collect_s=18.9 train_s=16.9 gpu_mem%=91 min_mean_util%=55 [gpu0=27% gpu1=0% gpu2=12% gpu3=37% gpu4=0% gpu5=38% gpu6=11% gpu7=0%] [update 40] steps/s= 908.8 decisions_total= 671744 eps_done= 3335 recent_reward= 9.733 pg=+0.1535 v=14.4100 ent=1.706 entq=-0.060 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=780s update_s=18.0 collect_s=18.0 train_s=17.0 gpu_mem%=91 min_mean_util%=55 [gpu0=25% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=11% gpu7=0%] [update 41] steps/s= 936.2 decisions_total= 688128 eps_done= 3376 recent_reward= 6.786 pg=+0.0639 v=67.6912 ent=1.153 entq=-0.337 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=798s update_s=17.5 collect_s=17.5 train_s=15.6 gpu_mem%=91 min_mean_util%=55 [gpu0=11% gpu1=35% gpu2=32% gpu3=26% gpu4=29% gpu5=32% gpu6=26% gpu7=26%] [update 42] steps/s= 883.8 decisions_total= 704512 eps_done= 3584 recent_reward= 6.805 pg=+0.1417 v=9.2507 ent=2.028 entq=-0.102 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=816s update_s=18.5 collect_s=18.5 train_s=18.2 gpu_mem%=91 min_mean_util%=55 [gpu0=100% gpu1=100% gpu2=100% gpu3=100% gpu4=100% gpu5=100% gpu6=100% gpu7=100%] [update 43] steps/s= 914.4 decisions_total= 720896 eps_done= 3588 recent_reward= 7.110 pg=+0.1606 v=36.4001 ent=0.769 entq=-0.068 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=834s update_s=17.9 collect_s=17.9 train_s=16.2 gpu_mem%=91 min_mean_util%=55 [gpu0=11% gpu1=0% gpu2=27% gpu3=0% gpu4=29% gpu5=24% gpu6=30% gpu7=14%] [update 44] steps/s= 859.7 decisions_total= 737280 eps_done= 3627 recent_reward= 7.153 pg=+0.1530 v=2.1996 ent=1.711 entq=-0.563 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=853s update_s=19.1 collect_s=19.0 train_s=17.7 gpu_mem%=91 min_mean_util%=55 [gpu0=19% gpu1=26% gpu2=27% gpu3=16% gpu4=24% gpu5=20% gpu6=0% gpu7=26%] [update 45] steps/s= 927.6 decisions_total= 753664 eps_done= 3840 recent_reward= 6.514 pg=+0.0658 v=276.4028 ent=2.131 entq=-0.025 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=871s update_s=17.7 collect_s=17.6 train_s=16.1 gpu_mem%=91 min_mean_util%=56 [gpu0=31% gpu1=9% gpu2=26% gpu3=27% gpu4=29% gpu5=34% gpu6=25% gpu7=31%] [update 46] steps/s= 920.1 decisions_total= 770048 eps_done= 3844 recent_reward= 6.320 pg=+0.4103 v=64.7149 ent=1.852 entq=-0.016 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=889s update_s=17.8 collect_s=17.8 train_s=16.6 gpu_mem%=91 min_mean_util%=56 [gpu0=27% gpu1=25% gpu2=27% gpu3=26% gpu4=25% gpu5=25% gpu6=16% gpu7=29%] [update 47] steps/s= 871.8 decisions_total= 786432 eps_done= 3881 recent_reward= 6.557 pg=+0.0727 v=10.1575 ent=1.921 entq=-0.035 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=908s update_s=18.8 collect_s=18.8 train_s=16.7 gpu_mem%=91 min_mean_util%=56 [gpu0=27% gpu1=23% gpu2=23% gpu3=23% gpu4=21% gpu5=21% gpu6=18% gpu7=25%] [update 48] steps/s= 815.5 decisions_total= 802816 eps_done= 4096 recent_reward= 7.071 pg=+0.1244 v=0.5447 ent=3.189 entq=-0.024 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=928s update_s=20.1 collect_s=20.0 train_s=17.4 gpu_mem%=91 min_mean_util%=56 [gpu0=28% gpu1=0% gpu2=0% gpu3=28% gpu4=0% gpu5=0% gpu6=0% gpu7=12%] [update 49] steps/s= 870.6 decisions_total= 819200 eps_done= 4099 recent_reward= 7.074 pg=+0.2330 v=15.2458 ent=3.855 entq=-0.152 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=946s update_s=18.8 collect_s=18.8 train_s=16.6 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=30% gpu2=26% gpu3=31% gpu4=15% gpu5=0% gpu6=25% gpu7=27%] [update 50] steps/s= 840.6 decisions_total= 835584 eps_done= 4131 recent_reward= 7.517 pg=+0.2130 v=3.4246 ent=1.255 entq=-0.427 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=966s update_s=19.5 collect_s=19.4 train_s=17.1 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=30% gpu2=0% gpu3=29% gpu4=26% gpu5=28% gpu6=0% gpu7=0%] [train] checkpoint saved: checkpoints/policy_update50.ot [update 51] steps/s= 834.8 decisions_total= 851968 eps_done= 4183 recent_reward= 7.263 pg=+0.1109 v=1.3679 ent=2.864 entq=-0.025 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=986s update_s=19.6 collect_s=19.6 train_s=16.9 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=24% gpu2=29% gpu3=20% gpu4=30% gpu5=27% gpu6=21% gpu7=29%] [update 52] steps/s= 840.5 decisions_total= 868352 eps_done= 4355 recent_reward= 6.752 pg=+0.1408 v=1.5845 ent=5.235 entq=-0.020 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1005s update_s=19.5 collect_s=19.4 train_s=17.5 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=25% gpu2=25% gpu3=0% gpu4=26% gpu5=25% gpu6=26% gpu7=0%] [update 53] steps/s= 859.3 decisions_total= 884736 eps_done= 4383 recent_reward= 7.466 pg=+0.4548 v=7.2336 ent=1.540 entq=-0.006 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1024s update_s=19.1 collect_s=19.0 train_s=17.2 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=27% gpu2=25% gpu3=27% gpu4=25% gpu5=26% gpu6=22% gpu7=11%] [update 54] steps/s= 854.0 decisions_total= 901120 eps_done= 4434 recent_reward= 6.841 pg=+0.0554 v=2.4454 ent=4.020 entq=-0.010 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1043s update_s=19.2 collect_s=19.1 train_s=18.3 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=13% gpu2=31% gpu3=33% gpu4=17% gpu5=31% gpu6=27% gpu7=24%] [update 55] steps/s= 891.1 decisions_total= 917504 eps_done= 4610 recent_reward= 7.128 pg=+0.2481 v=2.0204 ent=2.558 entq=-0.002 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1062s update_s=18.4 collect_s=18.3 train_s=16.9 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=13% gpu2=24% gpu3=27% gpu4=28% gpu5=26% gpu6=27% gpu7=23%] [update 56] steps/s= 836.8 decisions_total= 933888 eps_done= 4636 recent_reward= 7.561 pg=+0.0652 v=2.0617 ent=2.211 entq=-0.012 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1081s update_s=19.6 collect_s=19.5 train_s=18.1 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=0% gpu2=14% gpu3=34% gpu4=0% gpu5=0% gpu6=0% gpu7=31%] [update 57] steps/s= 865.6 decisions_total= 950272 eps_done= 4684 recent_reward= 9.464 pg=+0.0590 v=3.7349 ent=3.008 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1100s update_s=18.9 collect_s=18.9 train_s=16.6 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=31% gpu2=29% gpu3=28% gpu4=0% gpu5=0% gpu6=11% gpu7=28%] [update 58] steps/s= 875.4 decisions_total= 966656 eps_done= 4866 recent_reward= 7.369 pg=+0.0953 v=4.2191 ent=2.074 entq=-0.004 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1119s update_s=18.7 collect_s=18.7 train_s=17.3 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=0% gpu2=30% gpu3=30% gpu4=36% gpu5=22% gpu6=26% gpu7=23%] [update 59] steps/s= 876.7 decisions_total= 983040 eps_done= 4881 recent_reward= 7.282 pg=+0.1328 v=3.7055 ent=1.098 entq=-0.005 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1138s update_s=18.7 collect_s=18.6 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=17% gpu1=14% gpu2=16% gpu3=14% gpu4=25% gpu5=26% gpu6=11% gpu7=27%] [update 60] steps/s= 879.3 decisions_total= 999424 eps_done= 4944 recent_reward= 7.962 pg=+0.0570 v=1.5570 ent=2.222 entq=-0.009 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1156s update_s=18.6 collect_s=18.6 train_s=17.6 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=25%] [update 61] steps/s= 885.6 decisions_total= 1015808 eps_done= 5122 recent_reward= 8.817 pg=+0.1703 v=7.4907 ent=3.351 entq=-0.002 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1175s update_s=18.5 collect_s=18.5 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=0% gpu1=0% gpu2=0% gpu3=0% gpu4=26% gpu5=0% gpu6=0% gpu7=2%] [update 62] steps/s= 851.2 decisions_total= 1032192 eps_done= 5135 recent_reward= 9.150 pg=-0.0287 v=2.1823 ent=2.455 entq=-0.008 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1194s update_s=19.2 collect_s=19.2 train_s=18.0 gpu_mem%=92 min_mean_util%=56 [gpu0=24% gpu1=0% gpu2=25% gpu3=25% gpu4=10% gpu5=10% gpu6=0% gpu7=26%] [update 63] steps/s= 898.1 decisions_total= 1048576 eps_done= 5193 recent_reward= 7.844 pg=+0.1319 v=1.3819 ent=1.254 entq=-0.003 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1212s update_s=18.2 collect_s=18.2 train_s=17.8 gpu_mem%=92 min_mean_util%=56 [gpu0=21% gpu1=2% gpu2=27% gpu3=10% gpu4=14% gpu5=29% gpu6=34% gpu7=72%] [update 64] steps/s= 869.8 decisions_total= 1064960 eps_done= 5376 recent_reward= 6.625 pg=+0.0822 v=21.5305 ent=1.133 entq=-0.002 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1231s update_s=18.8 collect_s=18.8 train_s=18.3 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=28% gpu2=29% gpu3=21% gpu4=26% gpu5=25% gpu6=27% gpu7=26%] [update 65] steps/s= 889.7 decisions_total= 1081344 eps_done= 5388 recent_reward= 6.374 pg=-0.0455 v=4.8901 ent=0.593 entq=-0.003 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1250s update_s=18.4 collect_s=18.4 train_s=16.0 gpu_mem%=92 min_mean_util%=56 [gpu0=32% gpu1=29% gpu2=26% gpu3=25% gpu4=36% gpu5=26% gpu6=26% gpu7=25%] [update 66] steps/s= 862.7 decisions_total= 1097728 eps_done= 5444 recent_reward= 8.320 pg=+0.1025 v=2.9835 ent=0.512 entq=-0.016 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1269s update_s=19.0 collect_s=18.9 train_s=16.3 gpu_mem%=92 min_mean_util%=56 [gpu0=17% gpu1=26% gpu2=29% gpu3=25% gpu4=14% gpu5=29% gpu6=15% gpu7=0%] [update 67] steps/s= 897.1 decisions_total= 1114112 eps_done= 5632 recent_reward= 7.000 pg=+0.0951 v=2.2674 ent=0.537 entq=-0.001 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1287s update_s=18.3 collect_s=18.2 train_s=15.7 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=33% gpu2=30% gpu3=41% gpu4=27% gpu5=14% gpu6=39% gpu7=15%] [update 68] steps/s= 834.2 decisions_total= 1130496 eps_done= 5647 recent_reward= 6.900 pg=+0.0788 v=6.1331 ent=0.486 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1306s update_s=19.6 collect_s=19.6 train_s=17.1 gpu_mem%=92 min_mean_util%=56 [gpu0=14% gpu1=24% gpu2=11% gpu3=26% gpu4=27% gpu5=28% gpu6=0% gpu7=0%] [update 69] steps/s= 887.5 decisions_total= 1146880 eps_done= 5697 recent_reward= 7.900 pg=+0.0761 v=0.1837 ent=0.424 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1325s update_s=18.5 collect_s=18.4 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=31% gpu2=25% gpu3=24% gpu4=25% gpu5=23% gpu6=14% gpu7=25%] [update 70] steps/s= 871.7 decisions_total= 1163264 eps_done= 5888 recent_reward= 9.212 pg=+0.0077 v=4.6087 ent=1.136 entq=-0.002 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1344s update_s=18.8 collect_s=18.7 train_s=16.6 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=24% gpu2=0% gpu3=18% gpu4=27% gpu5=0% gpu6=26% gpu7=0%] [update 71] steps/s= 859.2 decisions_total= 1179648 eps_done= 5900 recent_reward= 8.463 pg=-0.0300 v=6.3514 ent=0.796 entq=-0.001 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1363s update_s=19.1 collect_s=19.0 train_s=16.9 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=24% gpu2=23% gpu3=0% gpu4=0% gpu5=25% gpu6=26% gpu7=0%] [update 72] steps/s= 897.5 decisions_total= 1196032 eps_done= 5952 recent_reward= 8.991 pg=+0.0648 v=0.8689 ent=0.767 entq=-0.001 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1381s update_s=18.3 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=0% gpu2=27% gpu3=27% gpu4=26% gpu5=0% gpu6=0% gpu7=0%] [update 73] steps/s= 887.7 decisions_total= 1212416 eps_done= 6144 recent_reward= 8.129 pg=+0.0347 v=1.3187 ent=0.980 entq=-0.000 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1399s update_s=18.5 collect_s=18.4 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=29% gpu2=27% gpu3=17% gpu4=13% gpu5=0% gpu6=22% gpu7=24%] [update 74] steps/s= 876.2 decisions_total= 1228800 eps_done= 6156 recent_reward= 7.538 pg=+0.0645 v=2.9938 ent=1.115 entq=-0.005 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1418s update_s=18.7 collect_s=18.7 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=30% gpu2=0% gpu3=28% gpu4=25% gpu5=30% gpu6=30% gpu7=21%] [update 75] steps/s= 866.8 decisions_total= 1245184 eps_done= 6201 recent_reward= 6.742 pg=+0.0255 v=0.4432 ent=0.276 entq=-0.003 ecoef=0.0099 stage=0 lr=3.00e-4 elapsed=1437s update_s=18.9 collect_s=18.8 train_s=16.3 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=28% gpu2=27% gpu3=0% gpu4=0% gpu5=0% gpu6=11% gpu7=11%] [train] checkpoint saved: checkpoints/policy_update75.ot [update 76] steps/s= 883.6 decisions_total= 1261568 eps_done= 6274 recent_reward= 8.364 pg=+0.0505 v=5.1202 ent=0.225 entq=-0.000 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1456s update_s=18.5 collect_s=18.5 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=25% gpu2=30% gpu3=28% gpu4=20% gpu5=26% gpu6=0% gpu7=26%] [update 77] steps/s= 877.8 decisions_total= 1277952 eps_done= 6410 recent_reward= 7.229 pg=+0.0123 v=2.2266 ent=0.338 entq=-0.000 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1474s update_s=18.7 collect_s=18.6 train_s=16.8 gpu_mem%=92 min_mean_util%=56 [gpu0=32% gpu1=11% gpu2=0% gpu3=12% gpu4=36% gpu5=0% gpu6=0% gpu7=0%] [update 78] steps/s= 879.2 decisions_total= 1294336 eps_done= 6450 recent_reward= 7.966 pg=+0.0006 v=0.3834 ent=0.418 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1493s update_s=18.6 collect_s=18.6 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=26% gpu2=0% gpu3=30% gpu4=28% gpu5=29% gpu6=0% gpu7=0%] [update 79] steps/s= 863.7 decisions_total= 1310720 eps_done= 6518 recent_reward= 11.603 pg=+0.0667 v=1.6787 ent=0.525 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1512s update_s=19.0 collect_s=18.9 train_s=16.2 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=9% gpu2=0% gpu3=29% gpu4=24% gpu5=12% gpu6=0% gpu7=0%] [update 80] steps/s= 868.2 decisions_total= 1327104 eps_done= 6666 recent_reward= 11.730 pg=-0.0604 v=1.2676 ent=0.753 entq=-0.005 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1531s update_s=18.9 collect_s=18.8 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=42% gpu2=26% gpu3=0% gpu4=13% gpu5=0% gpu6=0% gpu7=27%] [update 81] steps/s= 875.2 decisions_total= 1343488 eps_done= 6708 recent_reward= 10.121 pg=+0.0371 v=3.5656 ent=0.588 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1550s update_s=18.7 collect_s=18.7 train_s=16.6 gpu_mem%=92 min_mean_util%=56 [gpu0=14% gpu1=17% gpu2=21% gpu3=27% gpu4=26% gpu5=22% gpu6=26% gpu7=0%] [update 82] steps/s= 874.5 decisions_total= 1359872 eps_done= 6770 recent_reward= 8.533 pg=-0.0352 v=4.3367 ent=0.632 entq=-0.024 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1568s update_s=18.7 collect_s=18.7 train_s=16.2 gpu_mem%=92 min_mean_util%=56 [gpu0=22% gpu1=20% gpu2=33% gpu3=28% gpu4=20% gpu5=13% gpu6=24% gpu7=20%] [update 83] steps/s= 919.8 decisions_total= 1376256 eps_done= 6920 recent_reward= 9.205 pg=+0.0395 v=1.2532 ent=0.614 entq=-0.000 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1586s update_s=17.8 collect_s=17.8 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=32% gpu1=36% gpu2=26% gpu3=28% gpu4=29% gpu5=0% gpu6=25% gpu7=20%] [update 84] steps/s= 891.5 decisions_total= 1392640 eps_done= 6957 recent_reward= 8.582 pg=+0.0595 v=2.2908 ent=0.999 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1604s update_s=18.4 collect_s=18.3 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=19% gpu2=27% gpu3=26% gpu4=28% gpu5=23% gpu6=0% gpu7=26%] [update 85] steps/s= 889.4 decisions_total= 1409024 eps_done= 7023 recent_reward= 6.359 pg=-0.0235 v=0.2492 ent=0.201 entq=-0.000 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1623s update_s=18.4 collect_s=18.4 train_s=16.2 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=26% gpu2=27% gpu3=0% gpu4=26% gpu5=19% gpu6=0% gpu7=0%] [update 86] steps/s= 875.9 decisions_total= 1425408 eps_done= 7176 recent_reward= 6.734 pg=+0.0746 v=0.3159 ent=0.428 entq=-0.000 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1642s update_s=18.7 collect_s=18.7 train_s=17.0 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=11% gpu2=26% gpu3=26% gpu4=36% gpu5=0% gpu6=22% gpu7=0%] [update 87] steps/s= 848.9 decisions_total= 1441792 eps_done= 7211 recent_reward= 7.618 pg=+0.0034 v=1.3817 ent=0.303 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1661s update_s=19.3 collect_s=19.3 train_s=17.7 gpu_mem%=92 min_mean_util%=56 [gpu0=24% gpu1=29% gpu2=25% gpu3=16% gpu4=2% gpu5=0% gpu6=0% gpu7=0%] [update 88] steps/s= 868.1 decisions_total= 1458176 eps_done= 7270 recent_reward= 8.035 pg=+0.0772 v=0.5244 ent=0.273 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1680s update_s=18.9 collect_s=18.8 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=30% gpu2=19% gpu3=0% gpu4=26% gpu5=0% gpu6=0% gpu7=0%] [update 89] steps/s= 878.8 decisions_total= 1474560 eps_done= 7428 recent_reward= 7.923 pg=+0.0279 v=0.7294 ent=0.712 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1698s update_s=18.6 collect_s=18.6 train_s=15.7 gpu_mem%=92 min_mean_util%=56 [gpu0=12% gpu1=29% gpu2=0% gpu3=30% gpu4=30% gpu5=0% gpu6=0% gpu7=0%] [update 90] steps/s= 885.5 decisions_total= 1490944 eps_done= 7460 recent_reward= 8.973 pg=-0.0658 v=1.9296 ent=0.302 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1717s update_s=18.5 collect_s=18.5 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=28% gpu2=31% gpu3=24% gpu4=16% gpu5=13% gpu6=19% gpu7=26%] [update 91] steps/s= 863.6 decisions_total= 1507328 eps_done= 7521 recent_reward= 7.079 pg=+0.0176 v=0.1575 ent=0.219 entq=-0.001 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1736s update_s=19.0 collect_s=18.9 train_s=16.2 gpu_mem%=92 min_mean_util%=56 [gpu0=0% gpu1=29% gpu2=28% gpu3=0% gpu4=0% gpu5=26% gpu6=0% gpu7=0%] [update 92] steps/s= 883.6 decisions_total= 1523712 eps_done= 7684 recent_reward= 8.091 pg=-0.0775 v=1.3131 ent=0.512 entq=-0.002 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1754s update_s=18.5 collect_s=18.5 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=25% gpu2=28% gpu3=25% gpu4=0% gpu5=25% gpu6=26% gpu7=0%] [update 93] steps/s= 896.5 decisions_total= 1540096 eps_done= 7710 recent_reward= 7.384 pg=+0.0464 v=1.1568 ent=0.386 entq=-0.002 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1773s update_s=18.3 collect_s=18.2 train_s=16.2 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=24% gpu2=17% gpu3=12% gpu4=24% gpu5=25% gpu6=0% gpu7=24%] [update 94] steps/s= 884.2 decisions_total= 1556480 eps_done= 7769 recent_reward= 8.709 pg=-0.0230 v=0.1435 ent=0.239 entq=-0.009 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1791s update_s=18.5 collect_s=18.5 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=11% gpu2=27% gpu3=34% gpu4=25% gpu5=0% gpu6=26% gpu7=32%] [update 95] steps/s= 855.0 decisions_total= 1572864 eps_done= 7943 recent_reward= 8.105 pg=+0.0117 v=0.5599 ent=0.384 entq=-0.010 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1810s update_s=19.2 collect_s=19.1 train_s=16.7 gpu_mem%=92 min_mean_util%=56 [gpu0=0% gpu1=12% gpu2=0% gpu3=40% gpu4=0% gpu5=26% gpu6=23% gpu7=23%] [update 96] steps/s= 903.1 decisions_total= 1589248 eps_done= 7958 recent_reward= 7.452 pg=+0.0272 v=0.4786 ent=0.442 entq=-0.012 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1828s update_s=18.1 collect_s=18.1 train_s=16.9 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=7% gpu2=22% gpu3=23% gpu4=20% gpu5=0% gpu6=32% gpu7=29%] [update 97] steps/s= 886.5 decisions_total= 1605632 eps_done= 8021 recent_reward= 7.963 pg=+0.0298 v=0.3455 ent=0.282 entq=-0.009 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1847s update_s=18.5 collect_s=18.4 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=18% gpu1=0% gpu2=25% gpu3=29% gpu4=27% gpu5=0% gpu6=22% gpu7=13%] [update 98] steps/s= 880.1 decisions_total= 1622016 eps_done= 8199 recent_reward= 7.992 pg=+0.0065 v=1.0840 ent=0.364 entq=-0.012 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1866s update_s=18.6 collect_s=18.6 train_s=15.8 gpu_mem%=92 min_mean_util%=56 [gpu0=16% gpu1=28% gpu2=29% gpu3=18% gpu4=29% gpu5=0% gpu6=31% gpu7=0%] [update 99] steps/s= 890.1 decisions_total= 1638400 eps_done= 8213 recent_reward= 7.473 pg=+0.0225 v=1.7441 ent=0.235 entq=-0.018 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1884s update_s=18.4 collect_s=18.4 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=30% gpu2=26% gpu3=26% gpu4=21% gpu5=22% gpu6=26% gpu7=0%] [update 100] steps/s= 871.1 decisions_total= 1654784 eps_done= 8267 recent_reward= 8.127 pg=-0.0634 v=0.1477 ent=0.265 entq=-0.049 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1903s update_s=18.8 collect_s=18.8 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=25% gpu2=32% gpu3=29% gpu4=24% gpu5=30% gpu6=26% gpu7=12%] [train] checkpoint saved: checkpoints/policy_update100.ot [update 101] steps/s= 886.9 decisions_total= 1671168 eps_done= 8358 recent_reward= 9.329 pg=-0.0091 v=0.2003 ent=0.437 entq=-0.038 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1921s update_s=18.5 collect_s=18.4 train_s=16.0 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=26% gpu2=26% gpu3=33% gpu4=0% gpu5=0% gpu6=0% gpu7=11%] [update 102] steps/s= 875.3 decisions_total= 1687552 eps_done= 8465 recent_reward= 8.301 pg=+0.0129 v=0.5120 ent=0.373 entq=-0.035 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1940s update_s=18.7 collect_s=18.7 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=0% gpu2=28% gpu3=13% gpu4=0% gpu5=12% gpu6=12% gpu7=0%] [update 103] steps/s= 890.3 decisions_total= 1703936 eps_done= 8517 recent_reward= 8.089 pg=+0.0150 v=0.4572 ent=0.243 entq=-0.078 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1958s update_s=18.4 collect_s=18.4 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=36% gpu1=30% gpu2=26% gpu3=28% gpu4=25% gpu5=17% gpu6=0% gpu7=26%] [update 104] steps/s= 886.5 decisions_total= 1720320 eps_done= 8601 recent_reward= 10.746 pg=+0.0198 v=0.6576 ent=0.390 entq=-0.018 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1977s update_s=18.5 collect_s=18.4 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=32% gpu2=13% gpu3=28% gpu4=30% gpu5=29% gpu6=11% gpu7=30%] [update 105] steps/s= 893.5 decisions_total= 1736704 eps_done= 8719 recent_reward= 9.625 pg=-0.0222 v=0.9782 ent=0.319 entq=-0.017 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=1995s update_s=18.3 collect_s=18.3 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=23% gpu1=37% gpu2=26% gpu3=26% gpu4=0% gpu5=0% gpu6=26% gpu7=26%] [update 106] steps/s= 888.0 decisions_total= 1753088 eps_done= 8765 recent_reward= 9.360 pg=-0.0161 v=0.7943 ent=0.236 entq=-0.029 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2014s update_s=18.5 collect_s=18.4 train_s=15.8 gpu_mem%=92 min_mean_util%=56 [gpu0=34% gpu1=32% gpu2=29% gpu3=29% gpu4=0% gpu5=33% gpu6=29% gpu7=29%] [update 107] steps/s= 917.4 decisions_total= 1769472 eps_done= 8848 recent_reward= 7.888 pg=+0.0443 v=0.6490 ent=0.301 entq=-0.027 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2032s update_s=17.9 collect_s=17.8 train_s=15.7 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=34% gpu2=22% gpu3=26% gpu4=24% gpu5=27% gpu6=24% gpu7=22%] [update 108] steps/s= 881.7 decisions_total= 1785856 eps_done= 8974 recent_reward= 9.713 pg=+0.0010 v=1.6472 ent=0.346 entq=-0.045 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2050s update_s=18.6 collect_s=18.5 train_s=17.0 gpu_mem%=92 min_mean_util%=56 [gpu0=37% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=25% gpu7=0%] [update 109] steps/s= 891.0 decisions_total= 1802240 eps_done= 9015 recent_reward= 10.090 pg=+0.0794 v=0.9182 ent=0.300 entq=-0.049 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2069s update_s=18.4 collect_s=18.3 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=24% gpu1=27% gpu2=25% gpu3=25% gpu4=25% gpu5=28% gpu6=25% gpu7=26%] [update 110] steps/s= 902.0 decisions_total= 1818624 eps_done= 9093 recent_reward= 8.755 pg=+0.0387 v=0.5011 ent=0.364 entq=-0.066 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2087s update_s=18.2 collect_s=18.1 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=0% gpu2=23% gpu3=0% gpu4=27% gpu5=11% gpu6=25% gpu7=28%] [update 111] steps/s= 891.1 decisions_total= 1835008 eps_done= 9230 recent_reward= 8.743 pg=-0.0187 v=0.1789 ent=0.437 entq=-0.049 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2105s update_s=18.4 collect_s=18.3 train_s=16.0 gpu_mem%=92 min_mean_util%=56 [gpu0=20% gpu1=37% gpu2=19% gpu3=0% gpu4=26% gpu5=27% gpu6=14% gpu7=27%] [update 112] steps/s= 902.6 decisions_total= 1851392 eps_done= 9267 recent_reward= 9.799 pg=+0.0324 v=0.7585 ent=0.330 entq=-0.113 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2123s update_s=18.2 collect_s=18.1 train_s=15.8 gpu_mem%=92 min_mean_util%=56 [gpu0=16% gpu1=29% gpu2=35% gpu3=0% gpu4=30% gpu5=26% gpu6=29% gpu7=30%] [update 113] steps/s= 870.6 decisions_total= 1867776 eps_done= 9339 recent_reward= 9.091 pg=-0.0336 v=0.5311 ent=0.371 entq=-0.099 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2142s update_s=18.8 collect_s=18.8 train_s=16.7 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=0% gpu2=25% gpu3=19% gpu4=0% gpu5=30% gpu6=0% gpu7=0%] [update 114] steps/s= 891.5 decisions_total= 1884160 eps_done= 9482 recent_reward= 8.699 pg=-0.0660 v=0.3006 ent=0.541 entq=-0.069 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2160s update_s=18.4 collect_s=18.3 train_s=15.6 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=25% gpu2=13% gpu3=25% gpu4=31% gpu5=13% gpu6=29% gpu7=12%] [update 115] steps/s= 899.4 decisions_total= 1900544 eps_done= 9517 recent_reward= 9.582 pg=+0.0224 v=1.1650 ent=0.412 entq=-0.134 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2179s update_s=18.2 collect_s=18.2 train_s=16.8 gpu_mem%=92 min_mean_util%=56 [gpu0=24% gpu1=23% gpu2=27% gpu3=26% gpu4=18% gpu5=28% gpu6=27% gpu7=26%] [update 116] steps/s= 882.7 decisions_total= 1916928 eps_done= 9584 recent_reward= 10.474 pg=+0.0458 v=0.6485 ent=0.274 entq=-0.111 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2197s update_s=18.6 collect_s=18.5 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=33% gpu2=26% gpu3=25% gpu4=26% gpu5=0% gpu6=0% gpu7=25%] [update 117] steps/s= 897.0 decisions_total= 1933312 eps_done= 9738 recent_reward= 9.343 pg=-0.0055 v=0.6007 ent=0.485 entq=-0.154 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2215s update_s=18.3 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=35% gpu2=31% gpu3=0% gpu4=24% gpu5=26% gpu6=26% gpu7=25%] [update 118] steps/s= 889.4 decisions_total= 1949696 eps_done= 9765 recent_reward= 9.405 pg=+0.0594 v=0.5824 ent=0.379 entq=-0.165 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2234s update_s=18.4 collect_s=18.4 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=27% gpu2=25% gpu3=15% gpu4=12% gpu5=12% gpu6=0% gpu7=0%] [update 119] steps/s= 900.8 decisions_total= 1966080 eps_done= 9829 recent_reward= 10.488 pg=+0.0643 v=0.3039 ent=0.195 entq=-0.205 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2252s update_s=18.2 collect_s=18.1 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=28% gpu2=27% gpu3=26% gpu4=13% gpu5=0% gpu6=0% gpu7=24%] [update 120] steps/s= 872.9 decisions_total= 1982464 eps_done= 9993 recent_reward= 10.756 pg=-0.0133 v=0.7664 ent=0.309 entq=-0.138 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2271s update_s=18.8 collect_s=18.7 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=29% gpu2=27% gpu3=29% gpu4=0% gpu5=0% gpu6=0% gpu7=0%] [update 121] steps/s= 914.0 decisions_total= 1998848 eps_done=10013 recent_reward= 10.332 pg=-0.0416 v=0.7700 ent=0.310 entq=-0.214 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2289s update_s=17.9 collect_s=17.9 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=30% gpu2=20% gpu3=31% gpu4=25% gpu5=30% gpu6=26% gpu7=29%] [update 122] steps/s= 885.7 decisions_total= 2015232 eps_done=10078 recent_reward= 10.287 pg=-0.0269 v=0.3167 ent=0.362 entq=-0.188 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2307s update_s=18.5 collect_s=18.5 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=28% gpu2=0% gpu3=30% gpu4=26% gpu5=12% gpu6=0% gpu7=28%] [update 123] steps/s= 882.4 decisions_total= 2031616 eps_done=10248 recent_reward= 12.271 pg=-0.0377 v=0.5753 ent=0.583 entq=-0.113 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2326s update_s=18.6 collect_s=18.5 train_s=16.6 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=24% gpu2=25% gpu3=14% gpu4=12% gpu5=0% gpu6=11% gpu7=25%] [update 124] steps/s= 875.3 decisions_total= 2048000 eps_done=10267 recent_reward= 11.725 pg=-0.0278 v=0.7189 ent=0.413 entq=-0.237 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2345s update_s=18.7 collect_s=18.7 train_s=16.7 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=0% gpu2=26% gpu3=35% gpu4=25% gpu5=10% gpu6=10% gpu7=12%] [update 125] steps/s= 916.4 decisions_total= 2064384 eps_done=10322 recent_reward= 11.054 pg=+0.0192 v=1.6193 ent=0.256 entq=-0.167 ecoef=0.0098 stage=0 lr=3.00e-4 elapsed=2362s update_s=17.9 collect_s=17.8 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=37% gpu2=25% gpu3=30% gpu4=12% gpu5=26% gpu6=25% gpu7=28%] [train] checkpoint saved: checkpoints/policy_update125.ot [update 126] steps/s= 871.3 decisions_total= 2080768 eps_done=10416 recent_reward= 10.135 pg=+0.0483 v=1.5919 ent=0.245 entq=-0.151 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2381s update_s=18.8 collect_s=18.8 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=13% gpu1=23% gpu2=31% gpu3=29% gpu4=0% gpu5=30% gpu6=27% gpu7=0%] [update 127] steps/s= 870.7 decisions_total= 2097152 eps_done=10519 recent_reward= 11.071 pg=+0.0427 v=0.7848 ent=0.264 entq=-0.206 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2400s update_s=18.8 collect_s=18.8 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=12% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=0%] [update 128] steps/s= 866.4 decisions_total= 2113536 eps_done=10570 recent_reward= 9.975 pg=+0.0708 v=3.6043 ent=0.339 entq=-0.284 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2419s update_s=18.9 collect_s=18.9 train_s=16.1 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=0%] [update 129] steps/s= 908.3 decisions_total= 2129920 eps_done=10658 recent_reward= 10.754 pg=+0.0036 v=1.4271 ent=0.667 entq=-0.160 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2437s update_s=18.0 collect_s=18.0 train_s=15.8 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=30% gpu2=26% gpu3=30% gpu4=16% gpu5=26% gpu6=27% gpu7=26%] [update 130] steps/s= 879.9 decisions_total= 2146304 eps_done=10771 recent_reward= 12.104 pg=+0.0511 v=2.6346 ent=0.292 entq=-0.197 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2456s update_s=18.6 collect_s=18.6 train_s=16.4 gpu_mem%=92 min_mean_util%=56 [gpu0=28% gpu1=19% gpu2=18% gpu3=28% gpu4=25% gpu5=20% gpu6=25% gpu7=27%] [update 131] steps/s= 895.5 decisions_total= 2162688 eps_done=10817 recent_reward= 11.082 pg=+0.0345 v=1.9389 ent=0.294 entq=-0.226 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2474s update_s=18.3 collect_s=18.2 train_s=16.7 gpu_mem%=92 min_mean_util%=56 [gpu0=27% gpu1=24% gpu2=0% gpu3=12% gpu4=20% gpu5=23% gpu6=32% gpu7=27%] [update 132] steps/s= 915.7 decisions_total= 2179072 eps_done=10903 recent_reward= 12.500 pg=-0.0309 v=16.9644 ent=0.406 entq=-0.179 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2492s update_s=17.9 collect_s=17.8 train_s=16.5 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=35% gpu2=0% gpu3=26% gpu4=26% gpu5=29% gpu6=0% gpu7=26%] [update 133] steps/s= 920.7 decisions_total= 2195456 eps_done=11025 recent_reward= 12.793 pg=+0.0059 v=1.0619 ent=0.678 entq=-0.209 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2510s update_s=17.8 collect_s=17.7 train_s=15.6 gpu_mem%=92 min_mean_util%=56 [gpu0=25% gpu1=27% gpu2=22% gpu3=28% gpu4=27% gpu5=30% gpu6=25% gpu7=26%] [update 134] steps/s= 878.1 decisions_total= 2211840 eps_done=11067 recent_reward= 11.863 pg=+0.0306 v=1.7676 ent=0.404 entq=-0.284 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2528s update_s=18.7 collect_s=18.6 train_s=15.9 gpu_mem%=92 min_mean_util%=56 [gpu0=29% gpu1=33% gpu2=30% gpu3=26% gpu4=0% gpu5=13% gpu6=26% gpu7=28%] [update 135] steps/s= 896.3 decisions_total= 2228224 eps_done=11150 recent_reward= 11.661 pg=+0.0052 v=1.4625 ent=0.462 entq=-0.214 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2547s update_s=18.3 collect_s=18.2 train_s=16.8 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=0% gpu2=24% gpu3=34% gpu4=0% gpu5=0% gpu6=0% gpu7=25%] [update 136] steps/s= 909.6 decisions_total= 2244608 eps_done=11279 recent_reward= 12.640 pg=+0.0378 v=1.3370 ent=0.570 entq=-0.185 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2565s update_s=18.0 collect_s=18.0 train_s=15.6 gpu_mem%=92 min_mean_util%=56 [gpu0=31% gpu1=35% gpu2=0% gpu3=0% gpu4=0% gpu5=38% gpu6=28% gpu7=29%] [update 137] steps/s= 913.8 decisions_total= 2260992 eps_done=11318 recent_reward= 12.148 pg=+0.0065 v=2.1001 ent=0.427 entq=-0.244 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2583s update_s=17.9 collect_s=17.9 train_s=16.0 gpu_mem%=92 min_mean_util%=56 [gpu0=26% gpu1=28% gpu2=30% gpu3=19% gpu4=33% gpu5=30% gpu6=29% gpu7=30%] [update 138] steps/s= 889.3 decisions_total= 2277376 eps_done=11391 recent_reward= 10.451 pg=-0.0002 v=0.4428 ent=0.414 entq=-0.265 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2601s update_s=18.4 collect_s=18.4 train_s=16.0 gpu_mem%=92 min_mean_util%=56 [gpu0=34% gpu1=0% gpu2=0% gpu3=0% gpu4=29% gpu5=19% gpu6=0% gpu7=4%] [update 139] steps/s= 912.2 decisions_total= 2293760 eps_done=11531 recent_reward= 13.560 pg=+0.0119 v=1.0761 ent=0.403 entq=-0.179 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2619s update_s=18.0 collect_s=17.9 train_s=16.2 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=29% gpu2=13% gpu3=26% gpu4=27% gpu5=11% gpu6=25% gpu7=13%] [update 140] steps/s= 887.4 decisions_total= 2310144 eps_done=11568 recent_reward= 11.341 pg=+0.0020 v=1.1428 ent=0.308 entq=-0.343 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2637s update_s=18.5 collect_s=18.4 train_s=17.2 gpu_mem%=92 min_mean_util%=57 [gpu0=0% gpu1=0% gpu2=0% gpu3=11% gpu4=0% gpu5=25% gpu6=32% gpu7=34%] [update 141] steps/s= 900.2 decisions_total= 2326528 eps_done=11636 recent_reward= 11.019 pg=+0.0214 v=0.3894 ent=0.372 entq=-0.262 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2656s update_s=18.2 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=28% gpu1=26% gpu2=0% gpu3=32% gpu4=24% gpu5=33% gpu6=26% gpu7=27%] [update 142] steps/s= 903.5 decisions_total= 2342912 eps_done=11787 recent_reward= 11.683 pg=+0.0400 v=3.9972 ent=0.746 entq=-0.184 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2674s update_s=18.1 collect_s=18.1 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=28% gpu1=0% gpu2=17% gpu3=0% gpu4=0% gpu5=20% gpu6=5% gpu7=26%] [update 143] steps/s= 898.3 decisions_total= 2359296 eps_done=11815 recent_reward= 11.587 pg=-0.0173 v=1.0271 ent=0.617 entq=-0.229 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2692s update_s=18.2 collect_s=18.2 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=33% gpu1=0% gpu2=0% gpu3=13% gpu4=0% gpu5=29% gpu6=29% gpu7=17%] [update 144] steps/s= 894.8 decisions_total= 2375680 eps_done=11879 recent_reward= 12.280 pg=-0.0080 v=0.4445 ent=0.285 entq=-0.293 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2710s update_s=18.3 collect_s=18.3 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=12% gpu1=0% gpu2=36% gpu3=0% gpu4=13% gpu5=0% gpu6=0% gpu7=39%] [update 145] steps/s= 893.0 decisions_total= 2392064 eps_done=12042 recent_reward= 10.472 pg=-0.0104 v=0.8157 ent=0.585 entq=-0.249 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2729s update_s=18.3 collect_s=18.3 train_s=16.6 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=0% gpu2=37% gpu3=0% gpu4=11% gpu5=18% gpu6=1% gpu7=34%] [update 146] steps/s= 920.7 decisions_total= 2408448 eps_done=12062 recent_reward= 10.696 pg=-0.0928 v=1.5200 ent=0.233 entq=-0.226 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2746s update_s=17.8 collect_s=17.7 train_s=15.6 gpu_mem%=92 min_mean_util%=57 [gpu0=31% gpu1=23% gpu2=25% gpu3=29% gpu4=13% gpu5=29% gpu6=28% gpu7=22%] [update 147] steps/s= 901.7 decisions_total= 2424832 eps_done=12127 recent_reward= 11.644 pg=-0.0274 v=1.7125 ent=0.307 entq=-0.243 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2765s update_s=18.2 collect_s=18.1 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=14% gpu1=30% gpu2=12% gpu3=0% gpu4=35% gpu5=0% gpu6=29% gpu7=32%] [update 148] steps/s= 873.2 decisions_total= 2441216 eps_done=12297 recent_reward= 12.184 pg=+0.0216 v=2.6496 ent=0.436 entq=-0.231 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2783s update_s=18.8 collect_s=18.7 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=12%] [update 149] steps/s= 882.8 decisions_total= 2457600 eps_done=12316 recent_reward= 12.013 pg=-0.0067 v=2.0587 ent=0.464 entq=-0.267 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2802s update_s=18.6 collect_s=18.5 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=15% gpu1=17% gpu2=38% gpu3=28% gpu4=37% gpu5=26% gpu6=27% gpu7=26%] [update 150] steps/s= 898.9 decisions_total= 2473984 eps_done=12370 recent_reward= 12.309 pg=+0.0315 v=0.4984 ent=0.326 entq=-0.295 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2820s update_s=18.2 collect_s=18.2 train_s=15.8 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=0% gpu2=28% gpu3=13% gpu4=0% gpu5=28% gpu6=29% gpu7=29%] [train] checkpoint saved: checkpoints/policy_update150.ot [update 151] steps/s= 896.2 decisions_total= 2490368 eps_done=12466 recent_reward= 12.091 pg=+0.0408 v=0.9745 ent=0.642 entq=-0.242 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2838s update_s=18.3 collect_s=18.2 train_s=16.2 gpu_mem%=92 min_mean_util%=57 [gpu0=17% gpu1=32% gpu2=34% gpu3=33% gpu4=27% gpu5=27% gpu6=31% gpu7=22%] [update 152] steps/s= 896.0 decisions_total= 2506752 eps_done=12568 recent_reward= 11.571 pg=+0.0720 v=0.8441 ent=0.525 entq=-0.255 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2857s update_s=18.3 collect_s=18.2 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=37% gpu1=0% gpu2=13% gpu3=13% gpu4=32% gpu5=0% gpu6=34% gpu7=26%] [update 153] steps/s= 903.0 decisions_total= 2523136 eps_done=12619 recent_reward= 13.126 pg=+0.0370 v=0.7843 ent=0.472 entq=-0.331 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2875s update_s=18.1 collect_s=18.1 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=19% gpu1=29% gpu2=14% gpu3=0% gpu4=17% gpu5=10% gpu6=24% gpu7=26%] [update 154] steps/s= 873.6 decisions_total= 2539520 eps_done=12708 recent_reward= 12.371 pg=+0.0129 v=0.4588 ent=0.761 entq=-0.291 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2894s update_s=18.8 collect_s=18.7 train_s=16.4 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=0% gpu2=27% gpu3=0% gpu4=39% gpu5=27% gpu6=28% gpu7=30%] [update 155] steps/s= 883.6 decisions_total= 2555904 eps_done=12820 recent_reward= 10.995 pg=+0.0291 v=0.8400 ent=0.596 entq=-0.298 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2912s update_s=18.5 collect_s=18.5 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=41% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=0% gpu7=12%] [update 156] steps/s= 896.4 decisions_total= 2572288 eps_done=12866 recent_reward= 11.868 pg=+0.0003 v=0.6673 ent=0.469 entq=-0.355 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2930s update_s=18.3 collect_s=18.2 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=32% gpu1=0% gpu2=26% gpu3=0% gpu4=31% gpu5=29% gpu6=13% gpu7=30%] [update 157] steps/s= 898.0 decisions_total= 2588672 eps_done=12953 recent_reward= 12.281 pg=-0.0541 v=0.5386 ent=0.672 entq=-0.327 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2949s update_s=18.2 collect_s=18.2 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=29% gpu1=30% gpu2=0% gpu3=23% gpu4=28% gpu5=27% gpu6=27% gpu7=29%] [update 158] steps/s= 892.7 decisions_total= 2605056 eps_done=13074 recent_reward= 14.765 pg=-0.0079 v=1.3651 ent=0.350 entq=-0.310 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2967s update_s=18.4 collect_s=18.3 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=30% gpu2=0% gpu3=27% gpu4=0% gpu5=23% gpu6=14% gpu7=30%] [update 159] steps/s= 902.5 decisions_total= 2621440 eps_done=13116 recent_reward= 12.817 pg=-0.0316 v=0.9022 ent=0.453 entq=-0.361 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=2985s update_s=18.2 collect_s=18.1 train_s=15.5 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=0% gpu2=0% gpu3=13% gpu4=0% gpu5=0% gpu6=0% gpu7=28%] [update 160] steps/s= 901.6 decisions_total= 2637824 eps_done=13199 recent_reward= 12.246 pg=+0.0228 v=0.6447 ent=0.341 entq=-0.412 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3003s update_s=18.2 collect_s=18.1 train_s=16.9 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=23% gpu2=0% gpu3=21% gpu4=26% gpu5=25% gpu6=0% gpu7=26%] [update 161] steps/s= 934.7 decisions_total= 2654208 eps_done=13328 recent_reward= 12.273 pg=-0.0083 v=4.0213 ent=0.330 entq=-0.401 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3021s update_s=17.5 collect_s=17.5 train_s=16.6 gpu_mem%=92 min_mean_util%=57 [gpu0=12% gpu1=31% gpu2=28% gpu3=41% gpu4=33% gpu5=28% gpu6=28% gpu7=34%] [update 162] steps/s= 900.2 decisions_total= 2670592 eps_done=13367 recent_reward= 12.403 pg=+0.0492 v=1.3130 ent=0.424 entq=-0.445 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3039s update_s=18.2 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=39% gpu2=0% gpu3=29% gpu4=0% gpu5=0% gpu6=0% gpu7=27%] [update 163] steps/s= 900.9 decisions_total= 2686976 eps_done=13440 recent_reward= 12.116 pg=-0.0135 v=0.3100 ent=0.626 entq=-0.428 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3057s update_s=18.2 collect_s=18.1 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=2% gpu2=26% gpu3=26% gpu4=27% gpu5=24% gpu6=28% gpu7=26%] [update 164] steps/s= 890.6 decisions_total= 2703360 eps_done=13580 recent_reward= 13.358 pg=+0.0820 v=0.9293 ent=0.335 entq=-0.363 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3076s update_s=18.4 collect_s=18.4 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=29% gpu2=0% gpu3=26% gpu4=29% gpu5=26% gpu6=27% gpu7=25%] [update 165] steps/s= 899.5 decisions_total= 2719744 eps_done=13617 recent_reward= 13.111 pg=+0.0212 v=0.6901 ent=0.390 entq=-0.537 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3094s update_s=18.2 collect_s=18.2 train_s=16.2 gpu_mem%=92 min_mean_util%=57 [gpu0=23% gpu1=0% gpu2=0% gpu3=0% gpu4=25% gpu5=23% gpu6=11% gpu7=25%] [update 166] steps/s= 899.2 decisions_total= 2736128 eps_done=13685 recent_reward= 11.686 pg=+0.0668 v=0.6906 ent=0.310 entq=-0.413 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3112s update_s=18.2 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=22% gpu1=28% gpu2=26% gpu3=29% gpu4=20% gpu5=27% gpu6=29% gpu7=35%] [update 167] steps/s= 913.5 decisions_total= 2752512 eps_done=13835 recent_reward= 12.603 pg=-0.0064 v=0.7260 ent=0.402 entq=-0.532 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3130s update_s=17.9 collect_s=17.9 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=25% gpu2=27% gpu3=26% gpu4=27% gpu5=36% gpu6=26% gpu7=26%] [update 168] steps/s= 880.3 decisions_total= 2768896 eps_done=13863 recent_reward= 11.750 pg=-0.0076 v=1.1415 ent=0.461 entq=-0.608 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3149s update_s=18.6 collect_s=18.6 train_s=16.6 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=28% gpu2=29% gpu3=0% gpu4=25% gpu5=27% gpu6=27% gpu7=26%] [update 169] steps/s= 911.0 decisions_total= 2785280 eps_done=13928 recent_reward= 13.797 pg=+0.0019 v=0.6977 ent=0.284 entq=-0.545 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3167s update_s=18.0 collect_s=17.9 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=34% gpu2=25% gpu3=25% gpu4=27% gpu5=23% gpu6=29% gpu7=31%] [update 170] steps/s= 909.1 decisions_total= 2801664 eps_done=14090 recent_reward= 12.409 pg=-0.0018 v=0.8157 ent=0.379 entq=-0.508 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3185s update_s=18.0 collect_s=18.0 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=31% gpu2=26% gpu3=12% gpu4=23% gpu5=26% gpu6=25% gpu7=28%] [update 171] steps/s= 917.0 decisions_total= 2818048 eps_done=14110 recent_reward= 12.829 pg=+0.0069 v=1.7222 ent=0.435 entq=-0.607 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3203s update_s=17.9 collect_s=17.8 train_s=16.2 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=27% gpu2=28% gpu3=28% gpu4=29% gpu5=32% gpu6=29% gpu7=33%] [update 172] steps/s= 885.0 decisions_total= 2834432 eps_done=14178 recent_reward= 11.272 pg=+0.0134 v=0.9034 ent=0.460 entq=-0.624 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3221s update_s=18.5 collect_s=18.5 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=0% gpu1=0% gpu2=0% gpu3=0% gpu4=28% gpu5=0% gpu6=0% gpu7=0%] [update 173] steps/s= 919.5 decisions_total= 2850816 eps_done=14345 recent_reward= 12.761 pg=-0.0012 v=1.0982 ent=0.500 entq=-0.453 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3239s update_s=17.8 collect_s=17.8 train_s=15.6 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=22% gpu2=27% gpu3=26% gpu4=24% gpu5=25% gpu6=17% gpu7=23%] [update 174] steps/s= 898.5 decisions_total= 2867200 eps_done=14364 recent_reward= 12.563 pg=-0.0421 v=1.6822 ent=0.335 entq=-0.531 ecoef=0.0097 stage=0 lr=3.00e-4 elapsed=3257s update_s=18.2 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=3% gpu1=24% gpu2=0% gpu3=36% gpu4=12% gpu5=0% gpu6=0% gpu7=28%] [update 175] steps/s= 921.5 decisions_total= 2883584 eps_done=14421 recent_reward= 12.300 pg=-0.0188 v=0.2347 ent=0.220 entq=-0.454 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3275s update_s=17.8 collect_s=17.7 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=17% gpu1=28% gpu2=25% gpu3=24% gpu4=24% gpu5=26% gpu6=15% gpu7=25%] [train] checkpoint saved: checkpoints/policy_update175.ot [update 176] steps/s= 919.5 decisions_total= 2899968 eps_done=14516 recent_reward= 13.818 pg=-0.0269 v=1.4778 ent=0.312 entq=-0.696 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3293s update_s=17.8 collect_s=17.8 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=32% gpu2=10% gpu3=27% gpu4=29% gpu5=28% gpu6=0% gpu7=25%] [update 177] steps/s= 907.1 decisions_total= 2916352 eps_done=14616 recent_reward= 13.286 pg=+0.0414 v=1.3076 ent=0.234 entq=-0.544 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3311s update_s=18.1 collect_s=18.0 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=28% gpu1=26% gpu2=0% gpu3=32% gpu4=24% gpu5=0% gpu6=36% gpu7=25%] [update 178] steps/s= 904.8 decisions_total= 2932736 eps_done=14669 recent_reward= 13.530 pg=+0.0120 v=0.9107 ent=0.225 entq=-0.617 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3329s update_s=18.1 collect_s=18.1 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=31% gpu1=16% gpu2=32% gpu3=27% gpu4=32% gpu5=0% gpu6=21% gpu7=28%] [update 179] steps/s= 894.1 decisions_total= 2949120 eps_done=14759 recent_reward= 12.927 pg=+0.0069 v=0.6857 ent=0.229 entq=-0.672 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3347s update_s=18.3 collect_s=18.3 train_s=16.5 gpu_mem%=92 min_mean_util%=57 [gpu0=17% gpu1=27% gpu2=25% gpu3=24% gpu4=25% gpu5=28% gpu6=0% gpu7=28%] [update 180] steps/s= 904.3 decisions_total= 2965504 eps_done=14869 recent_reward= 13.611 pg=-0.0329 v=1.4909 ent=0.198 entq=-0.635 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3365s update_s=18.1 collect_s=18.1 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=26% gpu2=11% gpu3=11% gpu4=29% gpu5=32% gpu6=29% gpu7=0%] [update 181] steps/s= 900.4 decisions_total= 2981888 eps_done=14916 recent_reward= 12.328 pg=+0.0431 v=1.0894 ent=0.252 entq=-0.651 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3384s update_s=18.2 collect_s=18.1 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=25% gpu2=31% gpu3=29% gpu4=33% gpu5=18% gpu6=26% gpu7=23%] [update 182] steps/s= 887.6 decisions_total= 2998272 eps_done=15003 recent_reward= 12.524 pg=-0.0296 v=0.8555 ent=0.251 entq=-0.604 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3402s update_s=18.5 collect_s=18.4 train_s=16.4 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=28% gpu2=29% gpu3=27% gpu4=20% gpu5=27% gpu6=22% gpu7=26%] [update 183] steps/s= 914.0 decisions_total= 3014656 eps_done=15123 recent_reward= 13.126 pg=-0.0292 v=0.7570 ent=0.321 entq=-0.607 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3420s update_s=17.9 collect_s=17.9 train_s=15.8 gpu_mem%=92 min_mean_util%=57 [gpu0=22% gpu1=28% gpu2=30% gpu3=25% gpu4=15% gpu5=25% gpu6=26% gpu7=25%] [update 184] steps/s= 921.5 decisions_total= 3031040 eps_done=15166 recent_reward= 12.593 pg=-0.0600 v=1.0633 ent=0.301 entq=-0.637 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3438s update_s=17.8 collect_s=17.7 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=29% gpu1=16% gpu2=27% gpu3=0% gpu4=27% gpu5=26% gpu6=29% gpu7=27%] [update 185] steps/s= 920.1 decisions_total= 3047424 eps_done=15249 recent_reward= 11.695 pg=+0.0912 v=0.7731 ent=0.418 entq=-0.620 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3455s update_s=17.8 collect_s=17.8 train_s=17.4 gpu_mem%=92 min_mean_util%=57 [gpu0=22% gpu1=2% gpu2=0% gpu3=10% gpu4=30% gpu5=31% gpu6=23% gpu7=25%] [update 186] steps/s= 895.7 decisions_total= 3063808 eps_done=15377 recent_reward= 12.748 pg=-0.0150 v=1.3584 ent=0.306 entq=-0.609 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3474s update_s=18.3 collect_s=18.2 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=29% gpu1=19% gpu2=27% gpu3=15% gpu4=30% gpu5=26% gpu6=25% gpu7=14%] [update 187] steps/s= 870.5 decisions_total= 3080192 eps_done=15415 recent_reward= 13.153 pg=-0.0415 v=0.8847 ent=0.416 entq=-0.618 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3493s update_s=18.8 collect_s=18.8 train_s=16.7 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=26% gpu2=27% gpu3=27% gpu4=28% gpu5=30% gpu6=27% gpu7=28%] [update 188] steps/s= 928.0 decisions_total= 3096576 eps_done=15490 recent_reward= 12.543 pg=-0.0287 v=0.9971 ent=0.460 entq=-0.537 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3510s update_s=17.7 collect_s=17.6 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=22% gpu2=23% gpu3=28% gpu4=26% gpu5=26% gpu6=26% gpu7=25%] [update 189] steps/s= 913.3 decisions_total= 3112960 eps_done=15629 recent_reward= 13.360 pg=-0.0584 v=0.4916 ent=0.416 entq=-0.544 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3528s update_s=17.9 collect_s=17.9 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=28% gpu1=28% gpu2=26% gpu3=23% gpu4=30% gpu5=27% gpu6=27% gpu7=27%] [update 190] steps/s= 894.6 decisions_total= 3129344 eps_done=15665 recent_reward= 12.089 pg=+0.0870 v=0.9965 ent=0.429 entq=-0.536 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3547s update_s=18.3 collect_s=18.3 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=29% gpu1=20% gpu2=0% gpu3=0% gpu4=0% gpu5=29% gpu6=29% gpu7=29%] [update 191] steps/s= 922.6 decisions_total= 3145728 eps_done=15735 recent_reward= 12.636 pg=+0.0107 v=1.1614 ent=0.277 entq=-0.507 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3564s update_s=17.8 collect_s=17.7 train_s=15.5 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=33% gpu2=24% gpu3=26% gpu4=31% gpu5=27% gpu6=26% gpu7=25%] [update 192] steps/s= 892.0 decisions_total= 3162112 eps_done=15884 recent_reward= 12.777 pg=+0.0215 v=1.0759 ent=0.165 entq=-0.445 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3583s update_s=18.4 collect_s=18.3 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=38% gpu1=0% gpu2=0% gpu3=0% gpu4=0% gpu5=0% gpu6=26% gpu7=18%] [update 193] steps/s= 902.0 decisions_total= 3178496 eps_done=15911 recent_reward= 12.577 pg=+0.0422 v=2.0538 ent=0.283 entq=-0.520 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3601s update_s=18.2 collect_s=18.1 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=24% gpu1=28% gpu2=31% gpu3=24% gpu4=31% gpu5=28% gpu6=30% gpu7=25%] [update 194] steps/s= 897.1 decisions_total= 3194880 eps_done=15978 recent_reward= 13.988 pg=+0.0534 v=1.0551 ent=0.388 entq=-0.454 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3619s update_s=18.3 collect_s=18.2 train_s=15.7 gpu_mem%=92 min_mean_util%=57 [gpu0=31% gpu1=0% gpu2=16% gpu3=29% gpu4=30% gpu5=33% gpu6=28% gpu7=29%] [update 195] steps/s= 904.0 decisions_total= 3211264 eps_done=16139 recent_reward= 13.116 pg=-0.0405 v=1.4474 ent=0.457 entq=-0.467 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3637s update_s=18.1 collect_s=18.1 train_s=15.9 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=13% gpu2=28% gpu3=31% gpu4=36% gpu5=32% gpu6=29% gpu7=11%] [update 196] steps/s= 899.5 decisions_total= 3227648 eps_done=16159 recent_reward= 13.321 pg=+0.0271 v=1.0106 ent=0.174 entq=-0.319 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3655s update_s=18.2 collect_s=18.2 train_s=16.5 gpu_mem%=92 min_mean_util%=57 [gpu0=13% gpu1=31% gpu2=30% gpu3=30% gpu4=27% gpu5=20% gpu6=28% gpu7=25%] [update 197] steps/s= 899.2 decisions_total= 3244032 eps_done=16226 recent_reward= 14.090 pg=-0.0426 v=1.0650 ent=0.131 entq=-0.503 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3674s update_s=18.2 collect_s=18.2 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=17% gpu1=26% gpu2=37% gpu3=26% gpu4=0% gpu5=17% gpu6=25% gpu7=31%] [update 198] steps/s= 883.1 decisions_total= 3260416 eps_done=16394 recent_reward= 11.398 pg=-0.0504 v=1.9483 ent=0.367 entq=-0.498 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3692s update_s=18.6 collect_s=18.5 train_s=17.0 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=27% gpu2=0% gpu3=26% gpu4=27% gpu5=25% gpu6=26% gpu7=26%] [update 199] steps/s= 913.8 decisions_total= 3276800 eps_done=16413 recent_reward= 11.162 pg=-0.0559 v=2.2105 ent=0.441 entq=-0.461 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3710s update_s=17.9 collect_s=17.9 train_s=16.9 gpu_mem%=92 min_mean_util%=57 [gpu0=32% gpu1=25% gpu2=28% gpu3=27% gpu4=21% gpu5=24% gpu6=23% gpu7=26%] [update 200] steps/s= 905.3 decisions_total= 3293184 eps_done=16469 recent_reward= 12.163 pg=-0.0018 v=0.8120 ent=0.204 entq=-0.484 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3728s update_s=18.1 collect_s=18.1 train_s=15.5 gpu_mem%=92 min_mean_util%=57 [gpu0=36% gpu1=0% gpu2=32% gpu3=29% gpu4=31% gpu5=14% gpu6=12% gpu7=28%] [train] checkpoint saved: checkpoints/policy_update200.ot [update 201] steps/s= 902.7 decisions_total= 3309568 eps_done=16565 recent_reward= 12.997 pg=-0.0122 v=0.8939 ent=0.214 entq=-0.475 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3746s update_s=18.1 collect_s=18.1 train_s=16.3 gpu_mem%=92 min_mean_util%=57 [gpu0=0% gpu1=0% gpu2=25% gpu3=31% gpu4=11% gpu5=23% gpu6=26% gpu7=0%] [update 202] steps/s= 907.6 decisions_total= 3325952 eps_done=16665 recent_reward= 13.392 pg=+0.0383 v=0.5179 ent=0.262 entq=-0.419 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3764s update_s=18.1 collect_s=18.0 train_s=16.1 gpu_mem%=92 min_mean_util%=57 [gpu0=26% gpu1=13% gpu2=29% gpu3=18% gpu4=30% gpu5=36% gpu6=26% gpu7=25%] [update 203] steps/s= 901.3 decisions_total= 3342336 eps_done=16717 recent_reward= 13.606 pg=+0.0136 v=1.1313 ent=0.201 entq=-0.547 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3783s update_s=18.2 collect_s=18.1 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=20% gpu2=31% gpu3=27% gpu4=18% gpu5=29% gpu6=26% gpu7=19%] [update 204] steps/s= 893.4 decisions_total= 3358720 eps_done=16807 recent_reward= 13.655 pg=-0.0693 v=0.4960 ent=0.204 entq=-0.525 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3801s update_s=18.3 collect_s=18.3 train_s=15.8 gpu_mem%=92 min_mean_util%=57 [gpu0=30% gpu1=32% gpu2=20% gpu3=32% gpu4=34% gpu5=28% gpu6=27% gpu7=30%] [update 205] steps/s= 908.4 decisions_total= 3375104 eps_done=16917 recent_reward= 12.325 pg=-0.0135 v=1.0432 ent=0.272 entq=-0.476 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3819s update_s=18.0 collect_s=18.0 train_s=16.6 gpu_mem%=92 min_mean_util%=57 [gpu0=0% gpu1=6% gpu2=25% gpu3=24% gpu4=39% gpu5=0% gpu6=25% gpu7=29%] [update 206] steps/s= 897.5 decisions_total= 3391488 eps_done=16964 recent_reward= 12.865 pg=+0.0242 v=0.9418 ent=0.397 entq=-0.487 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3837s update_s=18.3 collect_s=18.2 train_s=17.0 gpu_mem%=92 min_mean_util%=57 [gpu0=0% gpu1=24% gpu2=25% gpu3=0% gpu4=0% gpu5=0% gpu6=11% gpu7=25%] [update 207] steps/s= 896.4 decisions_total= 3407872 eps_done=17051 recent_reward= 13.629 pg=-0.0272 v=0.7444 ent=0.341 entq=-0.422 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3856s update_s=18.3 collect_s=18.2 train_s=16.9 gpu_mem%=92 min_mean_util%=57 [gpu0=15% gpu1=27% gpu2=13% gpu3=0% gpu4=12% gpu5=26% gpu6=31% gpu7=29%] [update 208] steps/s= 899.1 decisions_total= 3424256 eps_done=17171 recent_reward= 13.650 pg=-0.0713 v=1.7546 ent=0.325 entq=-0.547 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3874s update_s=18.2 collect_s=18.2 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=27% gpu2=28% gpu3=26% gpu4=29% gpu5=31% gpu6=27% gpu7=21%] [update 209] steps/s= 888.7 decisions_total= 3440640 eps_done=17214 recent_reward= 14.003 pg=-0.0177 v=1.8682 ent=0.283 entq=-0.506 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3892s update_s=18.4 collect_s=18.4 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=27% gpu1=34% gpu2=24% gpu3=30% gpu4=25% gpu5=31% gpu6=25% gpu7=26%] [update 210] steps/s= 902.7 decisions_total= 3457024 eps_done=17297 recent_reward= 14.029 pg=-0.0253 v=0.5032 ent=0.350 entq=-0.481 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3910s update_s=18.2 collect_s=18.1 train_s=16.2 gpu_mem%=92 min_mean_util%=57 [gpu0=25% gpu1=26% gpu2=26% gpu3=23% gpu4=26% gpu5=26% gpu6=27% gpu7=0%] [update 211] steps/s= 906.8 decisions_total= 3473408 eps_done=17425 recent_reward= 12.726 pg=-0.0038 v=0.5749 ent=0.231 entq=-0.512 ecoef=0.0096 stage=0 lr=3.00e-4 elapsed=3928s update_s=18.1 collect_s=18.0 train_s=16.0 gpu_mem%=92 min_mean_util%=57 [gpu0=28% gpu1=12% gpu2=33% gpu3=26% gpu4=30% gpu5=0% gpu6=0% gpu7=28%]