Refactors setup.py for production-ready package build #32
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
Transforms the basic CUDA extension setup into a comprehensive package configuration with proper metadata, dependencies, and build optimization. Adds copyright header, comprehensive package metadata including author information, description, and PyPI classifiers for better discoverability. Implements dynamic CUDA architecture detection, version checking, and proper error handling for unsupported CUDA versions (requires 11.7+). Introduces NinjaBuildExtension with intelligent job allocation based on available CPU cores and memory to prevent OOM during compilation. Expands source file coverage to include comprehensive flash attention kernels for multiple head dimensions, data types, and attention variants (regular, causal, split). Adds build environment controls through environment variables for forced builds, CUDA skipping, and ABI compatibility.
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR refactors setup.py to support a production-ready package build with dynamic CUDA architecture detection, enhanced package metadata, and an optimized Ninja build extension.
- Introduces dynamic CUDA version checking and conditional compilation flags.
- Implements a custom NinjaBuildExtension that allocates build jobs based on CPU cores and available memory.
- Updates package configuration with comprehensive metadata and environment variable controls.
Comments suppressed due to low confidence (2)
setup.py:208
- [nitpick] Consider adding a brief comment to explain the purpose of the newly introduced flag -U__CUDA_NO_HALF2_OPERATORS__ for clarity and future maintainability.
"-U__CUDA_NO_HALF2_OPERATORS__",
setup.py:286
- [nitpick] Clarify in a comment that the conditional setting of cmdclass is intentional to allow a minimal installation when no CUDA extensions are built.
else {},
Comment on lines
+7
to
+10
| import re | ||
| import ast | ||
| import glob | ||
| import shutil |
There was a problem hiding this comment.
[nitpick] There are several imports (e.g., re, ast, glob, shutil) that do not appear to be used; consider removing any unused imports to improve code clarity.
Suggested change
| import re | |
| import ast | |
| import glob | |
| import shutil |
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.
Transforms the basic CUDA extension setup into a comprehensive package configuration with proper metadata, dependencies, and build optimization.
Adds copyright header, comprehensive package metadata including author information, description, and PyPI classifiers for better discoverability.
Implements dynamic CUDA architecture detection, version checking, and proper error handling for unsupported CUDA versions (requires 11.7+).
Introduces NinjaBuildExtension with intelligent job allocation based on available CPU cores and memory to prevent OOM during compilation.
Expands source file coverage to include comprehensive flash attention kernels for multiple head dimensions, data types, and attention variants (regular, causal, split).
Adds build environment controls through environment variables for forced builds, CUDA skipping, and ABI compatibility.