feat(grpo): mask env-flagged samples from policy loss - #3163
Merged
Conversation
macandro96
added
the
CI:Lfast
Contributor
Author
|
/ok to test c112b82 |
Signed-off-by: Anish Mahishi <amahishi@nvidia.com>
macandro96
force-pushed
the
amahishi/add-mask-sample-mr355
branch
from
July 13, 2026 15:21
c112b82 to
8112e34
Compare
Contributor
Author
|
/ok to test 8112e34 |
yfw
reviewed
Jul 13, 2026
Signed-off-by: Anish Mahishi <amahishi@nvidia.com>
Contributor
Author
|
/ok to test 27a0f20 |
yfw
approved these changes
Jul 13, 2026
yfw
enabled auto-merge (squash)
July 13, 2026 21:41
binhu-nv
pushed a commit
that referenced
this pull request
Jul 22, 2026
Signed-off-by: Anish Mahishi <amahishi@nvidia.com> Co-authored-by: Sugam Devare <sdevare@nvidia.com> (cherry picked from commit 4372db1)
binhu-nv
pushed a commit
that referenced
this pull request
Jul 31, 2026
Signed-off-by: Anish Mahishi <amahishi@nvidia.com> Co-authored-by: Sugam Devare <sdevare@nvidia.com> (cherry picked from commit 4372db1)
4 tasks
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
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 ?
Adds a per-sample gradient mask driven by the rollout environment. When
instance_config.mask_sampleisTrueon a rollout result, the sample'sloss_multiplieris zeroed out before the policy update — but the sample is kept for advantage computation (baseline / std).Cherry-pick commit here
Issues
List issues that this PR closes (syntax):
Usage
# Add a code snippet demonstrating how to use thisBefore your PR is "Ready for review"
Pre checks:
Additional Information