Delete rowwise_scaled_linear_cutlass kernels and APIs#3723
Merged
jerryzh168 merged 1 commit intomainfrom Jan 27, 2026
Merged
Conversation
Summary: TODO: Files Deleted: 1. CUDA source files: - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass.cuh - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass_s4s4.cu - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass_s8s4.cu - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/README.md 2. Python layout files: - torchao/prototype/dtypes/uintx/cutlass_int4_packed_layout.py - torchao/dtypes/uintx/cutlass_int4_packed_layout.py 3. Test and benchmark files: - test/test_ops_rowwise_scaled_linear_cutlass.py - benchmarks/benchmark_rowwise_scaled_linear_cutlass.py Files Modified: 1. torchao/ops.py - Removed lib.define for rowwise_scaled_linear_cutlass_s8s4 and rowwise_scaled_linear_cutlass_s4s4, and their function implementations 2. torchao/csrc/README.md - Removed reference to cutlass naming convention 3. torchao/dtypes/affine_quantized_tensor_ops.py - Removed imports and dispatch registrations for cutlass int4 4. torchao/prototype/dtypes/__init__.py and torchao/prototype/dtypes/uintx/__init__.py - Removed CutlassInt4PackedLayout exports 5. torchao/dtypes/__init__.py - Removed CutlassInt4PackedLayout import and export 6. torchao/quantization/__init__.py and torchao/quantization/quant_api.py - Removed CutlassInt4PackedLayout and Int4DynamicActivationInt4WeightConfig exports 7. torchao/prototype/quantization/quant_api.py - Removed Int4DynamicActivationInt4WeightConfig class and updated Int8DynamicActivationInt4WeightConfig to remove CutlassInt4PackedLayout handling 8. Test files (test/dtypes/test_affine_quantized.py, test/dtypes/test_uintx.py, test/quantization/test_quant_api.py, test/core/test_config.py, test/integration/test_vllm.py) - Removed all references to CutlassInt4PackedLayout and Int4DynamicActivationInt4WeightConfig 9. benchmarks/microbenchmarks/utils.py - Removed int8adq-int4w-symm quantization option Test Plan: Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3723
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 832e40e with merge base 79372e7 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
jerryzh168
added a commit
that referenced
this pull request
Jan 26, 2026
Summary: TODO: Files Deleted: 1. CUDA source files: - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass.cuh - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass_s4s4.cu - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/rowwise_scaled_linear_cutlass_s8s4.cu - torchao/csrc/cuda/rowwise_scaled_linear_cutlass/README.md 2. Python layout files: - torchao/prototype/dtypes/uintx/cutlass_int4_packed_layout.py - torchao/dtypes/uintx/cutlass_int4_packed_layout.py 3. Test and benchmark files: - test/test_ops_rowwise_scaled_linear_cutlass.py - benchmarks/benchmark_rowwise_scaled_linear_cutlass.py Files Modified: 1. torchao/ops.py - Removed lib.define for rowwise_scaled_linear_cutlass_s8s4 and rowwise_scaled_linear_cutlass_s4s4, and their function implementations 2. torchao/csrc/README.md - Removed reference to cutlass naming convention 3. torchao/dtypes/affine_quantized_tensor_ops.py - Removed imports and dispatch registrations for cutlass int4 4. torchao/prototype/dtypes/__init__.py and torchao/prototype/dtypes/uintx/__init__.py - Removed CutlassInt4PackedLayout exports 5. torchao/dtypes/__init__.py - Removed CutlassInt4PackedLayout import and export 6. torchao/quantization/__init__.py and torchao/quantization/quant_api.py - Removed CutlassInt4PackedLayout and Int4DynamicActivationInt4WeightConfig exports 7. torchao/prototype/quantization/quant_api.py - Removed Int4DynamicActivationInt4WeightConfig class and updated Int8DynamicActivationInt4WeightConfig to remove CutlassInt4PackedLayout handling 8. Test files (test/dtypes/test_affine_quantized.py, test/dtypes/test_uintx.py, test/quantization/test_quant_api.py, test/core/test_config.py, test/integration/test_vllm.py) - Removed all references to CutlassInt4PackedLayout and Int4DynamicActivationInt4WeightConfig 9. benchmarks/microbenchmarks/utils.py - Removed int8adq-int4w-symm quantization option Test Plan: Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 44b139f Pull Request resolved: #3723
meta-cla
bot
added
the
CLA Signed
jerryzh168
added
the
module: deprecation
andrewor14
added
the
module: bc-breaking
andrewor14
approved these changes
Jan 26, 2026
Contributor
andrewor14
left a comment
andrewor14
left a comment
There was a problem hiding this comment.
Looks great. Also cc @jcaip
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.
Stack from ghstack (oldest at bottom):
Summary:
Deleted
Int4DynamicActivationInt4WeightConfigandCutlassInt4PackedLayoutofInt8DynamicActivationInt4WeightConfigtop level API and related kernels since these are not usedBC breaking note:
We are removing
Int4DynamicActivationInt4WeightConfigandCutlassInt4PackedLayoutoption ofInt8DynamicActivationInt4WeightConfig0.15.0
0.16.0
Both configs are dropped. Please use torchao <= 0.15.0 to use these configs
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags: