State available in SampleBatch and ReplayBuffer #43
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Edilmo
force-pushed
the
edpalenc/replay-state
branch
from
962ad27
to
5d19589
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are these changes needed?
Currently, for recurrent/recursive models, the state is only available for policy evaluation during training, but it's not available in the SampleBatch hence is not accessible at the execution plans level which in turn means that is not present in the replay buffer. So, apex-like algorithms can not use memory models right now in RLlib.
Here we are making the very first step towards supporting memory for this kind of algorithms.
Related issue number
None
Checks
scripts/format.shto lint the changes in this PR.