Updated to new CUDA version#103
Open
ljn7 wants to merge 15 commits into
Open
Conversation
Pytorch gpu allocator
…rp-transducer) - Adapted and updated for compatibility and enhancements on my fork - Updated relevant documentation and options - Updated CMakeLists.txt to support CUDA and CMake version selection Updated CMakeLists.txt to support CUDA and CMake version detection/selection
Detect CUDA to on/off WITH_GPU by default
- Check CUDA availability before declaring it as project language - Fall back to CPU-only build when CUDA toolkit not found - Prevents "Failed to find nvcc" error on systems without CUDA - Maintains GPU support when CUDA is properly installed
…atform checks - Enabled robust detection of CUDA and platform compatibility - Improved build logic to support CUDA builds on Windows - Added fallback mechanisms and clearer error handling - Ensured smoother CPU-only builds without redundant checks - Reduced platform-specific pain points for easier maintenance
- Added robust detection of CUDA toolkit using CUDAToolkit and CUDA packages - Introduced WITH_GPU option to allow user-controlled GPU builds - Automatically fallback to CPU-only build if CUDA is unavailable or WITH_GPU=OFF - Improved status messages to reflect user intent and system CUDA availability - Added support for CUDA architecture detection and override via CMAKE_CUDA_ARCHITECTURES - Improved build summary output for clarity
ljn7
force-pushed
the
master
branch
2 times, most recently
from
1f14512 to
a6f8c22
Compare
- Uses cmake and make to build and run test_cpu - Targets only CPU environments for faster and simpler CI
- Updated for compatability and performance similar to b-flo's implementation - Removed tensorflow and mxnet bindings (might add as git-modules)
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
1 participant
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.
This PR updates the PyTorch binding and CMake build setup to support newer CUDA versions (e.g., CUDA 12.9).
Changes:
Updated pytorch_binding/setup.py to work with recent CUDA installations.
Added support for setting
CFLAGS,LDFLAGS, andCUDA_HOMEmanually when using newerCUDAdirectory structures.Updated README with troubleshooting instructions.
Ensured compatibility with Python 3.12 and PyTorch compiled for newer environments.