[Fix] Remove task-type from RLVR recipe, update RLVR image selection logic#6071
Open
ehsu3 wants to merge 2 commits into
Open
[Fix] Remove task-type from RLVR recipe, update RLVR image selection logic#6071ehsu3 wants to merge 2 commits into
ehsu3 wants to merge 2 commits into
Conversation
Collaborator
|
You need to take this up with the Hub team: |
mujtaba1747
reviewed
Jul 21, 2026
| return textwrap.dedent(recipe_match.group(1)).strip() | ||
| result = textwrap.dedent(recipe_match.group(1)).strip() | ||
|
|
||
| # Strip task_type: storm_rbs from RFT/RLVR recipes - including it causes service validation failures. |
Collaborator
There was a problem hiding this comment.
This code is RLVR speicfic but will run for all other training techniques too
Collaborator
There was a problem hiding this comment.
Can we check training type (rlvr) before doing this check
Collaborator
There was a problem hiding this comment.
Also this is very nova specific, might want to use is_nova flag too
Contributor
Author
There was a problem hiding this comment.
Yep, I can add that check!
mujtaba1747
approved these changes
Jul 22, 2026
mujtaba1747
self-requested a review
July 22, 2026 20:28
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description of changes: Fixes two issues causing RLVR/RFT jobs on HyperPod to fail with service validation errors:
storm_rbsfrom recipes (specifically from RFT jobs)._extract_recipe_from_helm_template(), striptask_type: storm_rbsvia regex. Only strips the storm_rbs value (used by HPCLI), and will preserve other task_types used by other recipes.SM-HP-RFT-V2-latest, but the correct RFT training image usesSM-HP-RFT-TRAIN-V2-latest. Without TRAIN in the tag, the wrong container is pulled._train_hyperpod(): If the resolved image containsSM-HP-RFT-but notTRAIN, replaceSM-HP-RFT-withSM-HP-RFT-TRAIN-.Testing:
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.