docs: document MXFP numerics (mxfp8/mxfp4/nvfp4) and hardware dispatch#324
Open
lokic233 wants to merge 1 commit into
Open
docs: document MXFP numerics (mxfp8/mxfp4/nvfp4) and hardware dispatch#324lokic233 wants to merge 1 commit into
lokic233 wants to merge 1 commit into
Conversation
Qwix supports mxfp8, mxfp4, and nvfp4 as user-facing weight/activation qtypes, but they are absent from the user documentation. This adds a docs page covering: - Block/tile sizes (32 for mxfp8/mxfp4, 16 for nvfp4) - Scale format (e8m0fnu vs e4m3fn) - Hardware dispatch table (GPU fused via scaled_matmul; TPU/CPU native fp8 dot_general fallback) - Performance note about activation-quant overhead No code change — docs only.
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Qwix supports
mxfp8,mxfp4, andnvfp4as user-facingweight_qtype/act_qtype(seenumerics.py,qarray.pytile_size=32/16, and the PTQ/QT provider tests) but they are entirely absent from the user docs (docs/source/,README.md).This adds a
docs/source/mxfp.mdpage documenting:float8_e8m0fnuvsfloat8_e4m3fn)jax.nn.scaled_matmul; TPU/CPU fall through to native tiled fp8dot_general(_fast_dot_general), which is correct and MXU-acceleratedNo code change — docs only.