New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Adapt algo to 910B #156

Merged

WilfChen merged 1 commit into mindspore-lab:master from MashiroChen:master

Dec 11, 2023

Collaborator

MashiroChen commented Dec 11, 2023 •

edited

Loading

减少maddpg不必要控制流，如将learner部分从actor的循环中移出，删除if done，更新间隔为1的softupdate。性能提升明显从6.6ms/step -> 4.8ms/step
PPO算法修改环境并行线程数，ARM下30个线程性能较差更改为5。GE下kernel by kernel性能差于子图下沉端到端，修改为子图下沉执行。两项优化性能提升明显。环境线程修改性能从8.4ms/step->7.2ms/step
MAPPO适配910B

MashiroChen force-pushed the master branch 3 times, most recently from 861f335 to 0f16abb Compare

December 11, 2023 02:47


          Adapt algo to 910B

9525fc7

MashiroChen force-pushed the master branch from 0f16abb to 9525fc7 Compare

December 11, 2023 02:59

WilfChen merged commit 8598546 into mindspore-lab:master

0 of 3 checks passed

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet