Accept some patterns with Cutlass scheduler if option enabled #5077
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
…tlassScheduler classes with basic infrastructure
…eduler - implementing JIT compilation framework
- Reset CutlassCompiledKernel to standalone class (not inheriting from CompiledKernel) - Remove NVRTC compilation option, focus on nvcc system call compilation - Implement nvcc compilation pipeline with temporary directories and dlopen - Generate exact same kernel code as static nvfp4_scaled_mm.cu - Add proper error handling and compilation output capture - Update Cutlass scheduler to only accept ScaledMmaOp (already implemented) - Add CursorPlan.md documenting the implementation approach - Successfully builds and ready for testing with scaled_mm tests
- Added test_cutlass_executor.cpp with FusionExecutorCache tests - Updated test_cutlass_scheduler.cpp to use CutlassExecutor directly - Created cutlass_executor_example.py demonstrating nvfp4 scaled matmul - Added test_scaled_mm.cpp for comprehensive scaled_mm testing - Updated CMakeLists.txt to include new test files - All tests are building successfully and CutlassExecutor integration is working - Current issue: nvcc compilation failing due to missing cutlass_utils.h include path
This will need to be improved for actual automation
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Outdated suggestions cannot be applied.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.
C++ test passes, but running
NVFUSER_ENABLE=cutlass_scheduler pytest tests/python/test_narrow_precision.py -vsshows thattest_scaled_mmfails with a link error 🤕