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.

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

Prev_action, prev_reward not passed to rollout #4573

Closed
eugenevinitsky opened this issue Apr 6, 2019 · 2 comments
Closed

Prev_action, prev_reward not passed to rollout #4573

eugenevinitsky opened this issue Apr 6, 2019 · 2 comments

Comments

@eugenevinitsky
Copy link
Contributor

System information

  • OS Platform and Distribution (e.g., Linux Ubuntu 16.04): n/a
  • Ray installed from (source or binary): n/a
  • Ray version: n/a
  • Python version: n/a
  • Exact command to reproduce:

Describe the problem

rollout.py does not keep track of previous actions and previous observed rewards and pass them to compute_action.py. This obviously is an issue if the policy was trained with prev. observed rewards and actions.

Source code / logs

@ericl
Copy link
Contributor

ericl commented Apr 7, 2019

Ah yeah @vladfi1 has a fix here: #4565

@eugenevinitsky
Copy link
Contributor Author

Cool, resolved!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants