Repository navigation
Fix FP8 MLA KV scale propagation - #34765
Closed
deepankerk wants to merge 2 commits into
Closed
deepankerk wants to merge 2 commits into
deepankerk wants to merge 2 commits into
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Motivation
Calibrated FP8 MLA KV cache values must use the same scale convention when
they are written and read. Some MLA and DSA paths wrote the cache with unit
scale, restored only the K scale in BMM1, and omitted the V scale in BMM2.
That makes non-unit calibration inconsistent across the attention operation.
For the existing SGLang scale convention,
k_scaleandv_scalearedescales. Cache writes therefore multiply by the reciprocal descale, BMM1
applies the K descale with the attention softmax scale, and BMM2 applies the V
descale.
MLA stores one compressed latent row that is consumed as K in BMM1 and V in
BMM2. The K and V descales for that shared row must therefore agree.
Modifications
writer.
descales instead of silently using an incoherent scale contract.
required read scales.
This uses the existing
RadixAttention.k_scale,v_scale,k_scale_float, andv_scale_floatattributes. It does not introduce a newmodel-specific scale format.
Accuracy Tests
The included synthetic tests verify:
quantization.
Speed Tests and Profiling
This is a scale-correctness change. No performance claim is made in this PR.
Validation
The new synthetic tests are registered for the relevant upstream CI suites.
Checklist
CI States
Latest PR Test (Base): ❌ Run #31745650167
Latest PR Test (Extra): ❌ Run #31745649743