Make various updates and fixes:

- Add support for legacy CUDA versions; now compatible with CUDA 12.3 and newer
- Add support for NVRTC compilation
- Other fixes and code refactoring

install.sh

@@ -3,7 +3,7 @@ original_dir=$(pwd)
 script_dir=$(realpath "$(dirname "$0")")
 cd "$script_dir"
 
-# Remove old dist file, build, and install
+# Remove old dist file, build files, and install
 rm -rf build dist
 rm -rf *.egg-info
 python setup.py bdist_wheel