[GENERAL] Merged v1.0alpha into master. Added features are:

- A100 support via mma.16816
- Thread swizzling for conflict-free shared memory accesses without
padding
- Complete overhaul of the LLVM code generation in
codegen/selection/generator.cc to remove overengineering
- Added debugging capabilities in the Python binding
- Compilation error for kernels that spill
This commit is contained in:
Philippe Tillet
2021-01-11 19:20:34 -05:00
parent c0bc7ed8b0
commit 083bbd1e8d
75 changed files with 2688 additions and 4512 deletions

View File

@@ -111,7 +111,7 @@ setup(
author_email='ptillet@g.harvard.edu',
description='A language and compiler for custom Deep Learning operations',
long_description='',
packages=['triton', 'triton/_C', 'triton/ops'],
packages=['triton', 'triton/_C'],
install_requires=['numpy', 'torch', 'sympy'],
package_data={'': data},
ext_modules=[CMakeExtension('triton', 'triton/_C/')],