Bump flashinfer to nightly-v0.6.15-20260724 - #192
Draft
nehashah-eng wants to merge 2 commits into
Draft
Conversation
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
…outing patch The GLM-5.2 routing renorm 0/0 NaN bug this patch worked around (sigmoid underflow -> 0/0 in ScaledSumNormalizePostprocess) was fixed independently upstream two days later in flashinfer-ai/flashinfer#3803, merged on top of the block-per-token routing refactor (flashinfer-ai/flashinfer#3751). Both the mSumEpsilon=1e-20f fix and the lossy warp-path score reconstruction this patch addressed are already present in nightly-v0.6.15-20260724, so applying our patch on top just conflicts (as seen in CI: "patch failed: csrc/trtllm_fused_moe_runner.cu:76").
|
@nehashah-eng Build complete, success: https://github.com/coreweave/ml-containers/actions/runs/30131558781 |
|
@nehashah-eng Build complete, success: https://github.com/coreweave/ml-containers/actions/runs/30131558781 |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
flashinfer-commitin.github/configurations/vllm-tensorizer.yml(both cuda13.2.1 and cuda12.9.1 matrix entries) fromv0.6.14tonightly-v0.6.15-20260724.Test plan