Rename Sparse2x4CUTLASSFloat8Tensor to Float8Sparse2x4_2DData2DMetadataTensor (#4343)#4343
Open
bbeckca wants to merge 2 commits intopytorch:mainfrom
Open
Rename Sparse2x4CUTLASSFloat8Tensor to Float8Sparse2x4_2DData2DMetadataTensor (#4343)#4343bbeckca wants to merge 2 commits intopytorch:mainfrom
bbeckca wants to merge 2 commits intopytorch:mainfrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4343
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ❌ 2 New FailuresAs of commit 92e86c5 with merge base 6529fca ( NEW FAILURES - The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
jerryzh168
reviewed
Apr 27, 2026
Contributor
There was a problem hiding this comment.
can we just remove this file?
Contributor
There was a problem hiding this comment.
or print a warning here and direct to then new path
Contributor
Author
There was a problem hiding this comment.
Sounds good. Will redirect using an alias and delete this file.
jerryzh168
reviewed
Apr 27, 2026
Contributor
There was a problem hiding this comment.
is this an error? why do we have 1d_data_1d_metadata tensor here
jerryzh168
reviewed
Apr 27, 2026
|
|
||
| This packing format will dispatch to torch._cslt_sparse_mm for matmul, with per-tensor scaling passed as alpha. | ||
| """ | ||
| SPARSE_1D_DATA_1D_METADATA = "sparse_1d_data_1d_metadata" |
Contributor
There was a problem hiding this comment.
I thought this is added in a separate PR?
Contributor
There was a problem hiding this comment.
oh it's probably because of the PR is exported from stacked diffs
Contributor
Author
There was a problem hiding this comment.
I think you're right. Will publish from OSS next time to avoid this.
bbeckca
force-pushed
the
export-D102374347
branch
from
39bd83f to
b1bcda2
Compare
bbeckca
force-pushed
the
export-D102374347
branch
from
b1bcda2 to
bcd204e
Compare
|
@bbeckca has exported this pull request. If you are a Meta employee, you can view the originating Diff in D102374347. |
Contributor
Author
|
@pytorchbot label "module: not user facing" |
pytorch-bot
Bot
added
the
module: not user facing
meta-codesync
Bot
changed the title
bbeckca
added a commit
to bbeckca/ao
that referenced
this pull request
Apr 28, 2026
…taTensor (pytorch#4343) Summary: Rename the CUTLASS float8 sparse tensor class to describe the memory layout: - Class: Sparse2x4CUTLASSFloat8Tensor → Float8Sparse2x4_2DData2DMetadataTensor - Enum: SPARSE_CUTLASS → SPARSE_2D_DATA_2D_METADATA (old value kept for backward compatibility) The old identifiers to Sparse2x4CUTLASSFloat8Tensor will remain importable using backward compatible aliases. Reviewed By: RandySheriff Differential Revision: D102374347
bbeckca
force-pushed
the
export-D102374347
branch
from
bcd204e to
e452c2c
Compare
bbeckca
added a commit
to bbeckca/ao
that referenced
this pull request
Apr 28, 2026
…taTensor (pytorch#4343) Summary: Rename the CUTLASS float8 sparse tensor class to describe the memory layout: - Class: Sparse2x4CUTLASSFloat8Tensor → Float8Sparse2x4_2DData2DMetadataTensor - Enum: SPARSE_CUTLASS → SPARSE_2D_DATA_2D_METADATA (old value kept for backward compatibility) The old identifiers to Sparse2x4CUTLASSFloat8Tensor will remain importable using backward compatible aliases. Reviewed By: RandySheriff Differential Revision: D102374347
bbeckca
force-pushed
the
export-D102374347
branch
from
e452c2c to
f0135d7
Compare
Summary: X-link: pytorch/pytorch#180312 What: Adding a new tensor subclass for FP8 2:4 sparsity via hipSPARSELt (ROCm only). Packs compressed values + metadata into a single tensor with `_cslt_compress` and dispatches through `_cslt_sparse_mm` with `A_scale * B_scale` as `alpha`. Why: This hipSPARSELt path differs enough in packing and kernel routing from CUTLASS to warrant a dedicated path. Reference: https://rocm.blogs.amd.com/artificial-intelligence/introduce_hipsparselt/README.html Reviewed By: RandySheriff Differential Revision: D100640267
…taTensor (pytorch#4343) Summary: Rename the CUTLASS float8 sparse tensor class to describe the memory layout: - Class: Sparse2x4CUTLASSFloat8Tensor → Float8Sparse2x4_2DData2DMetadataTensor - Enum: SPARSE_CUTLASS → SPARSE_2D_DATA_2D_METADATA (old value kept for backward compatibility) The old identifiers to Sparse2x4CUTLASSFloat8Tensor will remain importable using backward compatible aliases. Reviewed By: RandySheriff Differential Revision: D102374347
bbeckca
force-pushed
the
export-D102374347
branch
from
f0135d7 to
92e86c5
Compare
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.
Summary:
Rename the CUTLASS float8 sparse tensor class to describe the memory layout:
The old identifiers to Sparse2x4CUTLASSFloat8Tensor will remain importable using backward compatible aliases.
Reviewed By: RandySheriff
Differential Revision: D102374347