Skip to content
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

commit working A2C vec env Pong #315

Merged
merged 7 commits into from
Apr 29, 2019
Merged

commit working A2C vec env Pong #315

merged 7 commits into from
Apr 29, 2019

Conversation

kengz
Copy link
Owner

@kengz kengz commented Apr 29, 2019 •
edited
Loading

commit working a2c Pong code

  • correct advantage and returns calculations in math util
  • correct gradient computation and loss functions
  • uses the fixed VecFrameStack state: fixed the state mutation bug in fix VecFrameStack state mutation bug #313
  • massive speedup: speed at over 1000 FPS (seen 1200) vs 150 FPS previously on 16 env and K80 GPU.

generalizing

  • fix remaining SIL issues. now tests pass
  • fix small PPO next_v_pred concat issue. Now runnable, results pending
  • remove heavy debug log to speed up env

This was referenced Apr 29, 2019
Closed
Closed
Merged
@kengz kengz merged commit 19cabed into v4-dev Apr 29, 2019
@kengz kengz deleted the a2c-rets branch April 29, 2019 15:31
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@kengz