fix(inference): swap known-issue precisions back (GB300 FP8, MI355X FP4)#442
Merged
Merged
Conversation
#441 applied the precision swap to the wrong entries. The GB300 Dynamo TRT MTP accuracy issue (NVIDIA/srt-slurm#51) affects FP8, and the MI355X MoRI SGLang MTP issue (sgl-project/sglang#27194) affects FP4. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
#441 applied the precision swap to the wrong entries — it set GB300 Dynamo TRT MTP to FP4 and MI355X MoRI SGLang MTP to FP8, which is backwards. This PR sets the correct precisions:
Testing
known-issues.test.ts,knownIssueAnnotations.test.ts, andcsv-export.test.ts; tests updated to assert the corrected precisions.KNOWN_CONFIG_ISSUESregistry — no matching-logic or rendering changes, so the official and?unofficialrun=overlay annotation paths are unaffected.🤖 Generated with Claude Code
Note
Low Risk
Registry-only precision strings in
KNOWN_CONFIG_ISSUESwith no matching or UI logic changes; only which benchmark points show known-issue warnings is affected.Overview
Corrects known-config issue precision filters that were swapped in a prior change: GB300 Dynamo TRT MTP now matches FP8 (NVIDIA/srt-slurm#51), and MI355X MoRI SGLang MTP matches FP4 (sgl-project/sglang#27194).
known-issues.test.tsis updated so positive/negative match cases and the deduplication scenario use those precisions. Matching, chart warnings, and CSV notes still use the samematchKnownConfigIssueslogic—only which(hwKey, precision)pairs trigger warnings changes.Reviewed by Cursor Bugbot for commit 7afd522. Bugbot is set up for automated code reviews on this repo. Configure here.