-
Notifications
You must be signed in to change notification settings - Fork 7.1k
[rllib] Add async remote workers #4253
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.
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
Conversation
|
@bjg2 this unifies the vectorized and multi-agent code paths for remote envs, and adds support for asynchronous execution. Let me know if I missed anything. |
|
Test FAILed. |
|
Test FAILed. |
|
Test FAILed. |
Looks good! I'm just not sure naming nailed it. Word Vector is omitted and Async is in there even though it might be sync (it is an argument). Maybe RemoveBaseEnv or something like that? |
| return obs | ||
|
|
||
|
|
||
| @ray.remote(num_cpus=0) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
wouldn't be surprised if this requires GPU resources in the future (no need to address this rn)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah, I think so too.
|
Test FAILed. |
What do these changes do?
Add a variant of
remote_worker_envs,async_remote_worker_envs, that does dynamic batching as envs complete. It also supports multi-agent.Related issue number
Closes #4235