refacotr(sunjx): refactor advantage calculation logic #16
Conversation
This file contains hidden or 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
puyuan1996
requested changes
Jan 4, 2026
|
|
||
| class REINFORCECalculator(DefaultAdvantageCalculator): | ||
| """ | ||
| Standard REINFORCE calculator. |
Collaborator
There was a problem hiding this comment.
I'm not sure if this advantage calculation corresponds to REINFORCE or REINFORCE++. Could you please clarify?
| Default calculator with no reward preprocessing. | ||
| Used by methods like GAE and CPGD that don't require reward preprocessing. | ||
| """ |
Collaborator
There was a problem hiding this comment.
It looks like REINFORCECalculator also uses this base class. Please update the overview accordingly."
| """ | ||
| Group normalization calculator. | ||
| Normalizes rewards within each group and optionally filters degenerate cases. |
Collaborator
There was a problem hiding this comment.
Please add the corresponding paper reference for each implemented AdvantageCalculator.
| ) | ||
| ) | ||
| # Prepare helper functions for calculator | ||
| get_gae_fn = ( |
Collaborator
There was a problem hiding this comment.
get_cumulative_returns and get_advantages_and_returns should also be moved into their respective AdvantageCalculator classes.
| # CPGD Utility Functions | ||
| # ============================================================================ | ||
|
|
||
| def _get_cpgd_advantages_returns( |
Collaborator
There was a problem hiding this comment.
_get_cpgd_advantages_returns should also be moved into their respective AdvantageCalculator classes.
puyuan1996
added
the
refactor
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
New advantage_calculator.py Module
Overview
A new module implementing various advantage calculation methods for reinforcement learning algorithms.
Abstract Base Class
Implementations
Factory Function