Remove reward model codepath#374
Merged
SumanthRH merged 5 commits intoNovaSky-AI:mainfrom Oct 2, 2025
Merged
Conversation
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
erictang000
approved these changes
Oct 1, 2025
Collaborator
erictang000
left a comment
erictang000
left a comment
There was a problem hiding this comment.
lgtm, if you haven't already can you try an e2e gsm8k run with ppo just to make sure, since the critic code path was more tightly coupled with reward?
erictang000
pushed a commit
that referenced
this pull request
Oct 2, 2025
li-boxuan
pushed a commit
to li-boxuan/SkyRL
that referenced
this pull request
Nov 23, 2025
# What does this PR do? Should close NovaSky-AI#371 We've had an unused codepath for using an outcome reward model in the training loop for a while. This was primarly for RLHF use-cases that we don't target and can be removed. TODO: - [x] Cleanup `custom_rewards` / `orm_rewards` keys in the trainer - [x] Cleanup `RewardModel` logic - [x] Cleanup `normalize_reward` config - [x] E2E test with gsm8k (GRPO and PPO) --------- Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
li-boxuan
pushed a commit
to li-boxuan/SkyRL
that referenced
this pull request
Nov 23, 2025
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Should close #371
We've had an unused codepath for using an outcome reward model in the training loop for a while. This was primarly for RLHF use-cases that we don't target and can be removed.
TODO:
custom_rewards/orm_rewardskeys in the trainerRewardModellogicnormalize_rewardconfig