Add three platform #907
Add three platform #907
Conversation
There was a problem hiding this comment.
Pull request overview
This pull request adds support for three hardware platforms (NVIDIA, MetaX, and Moore Threads) by implementing five new operators: atanh, addcmul, cdist, binary_cross_entropy_with_logits, and reciprocal. The changes include operator implementations for multiple backends (CPU, NVIDIA/CUDA, MetaX, Moore/MUSA), test files, and Python bindings.
Key changes:
- Implementation of 5 new operators across multiple hardware platforms
- Test infrastructure for infiniop and infinicore
- Python bindings and API registration
- Platform-specific kernel implementations for CUDA-like architectures
Reviewed changes
Copilot reviewed 114 out of 114 changed files in this pull request and generated 7 comments.
Show a summary per file
| File | Description |
|---|---|
| test/infiniop/*.py | Python test files for new operators (reciprocal, cdist, binary_cross_entropy_with_logits, atanh, addcmul) |
| test/infiniop/libinfiniop/op_register.py | Registration of new operators in the test framework |
| test/infinicore/ops/*.py | Uncommented infinicore operator implementations |
| src/infiniop/ops/*/operator.cc | Main operator dispatch logic for all platforms |
| src/infiniop/ops//nvidia/.cu(h) | NVIDIA/CUDA kernel implementations |
| src/infiniop/ops//metax/.maca | MetaX platform implementations |
| src/infiniop/ops//moore/.mu | Moore Threads (MUSA) implementations |
| src/infiniop/ops//cpu/.cc | CPU implementations |
| src/infiniop-test/src/ops/*.cpp | C++ test infrastructure |
| src/infinicore/* | InfiniCore operator implementations and Python bindings |
| python/infinicore/ops/*.py | Python wrapper for reciprocal operator |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| @@ -0,0 +1,163 @@ | |||
| import torch | |||
| import ctypes | |||
There was a problem hiding this comment.
Module 'ctypes' is imported with both 'import' and 'import from'.
| @@ -0,0 +1,171 @@ | |||
| import torch | |||
| import ctypes | |||
There was a problem hiding this comment.
Module 'ctypes' is imported with both 'import' and 'import from'.
| @@ -0,0 +1,154 @@ | |||
| import torch | |||
| import ctypes | |||
There was a problem hiding this comment.
Module 'ctypes' is imported with both 'import' and 'import from'.
| @@ -0,0 +1,156 @@ | |||
| import torch | |||
| import ctypes | |||
There was a problem hiding this comment.
Module 'ctypes' is imported with both 'import' and 'import from'.
| @@ -0,0 +1,165 @@ | |||
| import torch | |||
| import ctypes | |||
There was a problem hiding this comment.
Module 'ctypes' is imported with both 'import' and 'import from'.
| @@ -0,0 +1,154 @@ | |||
| import torch | |||
| import ctypes | |||
| from ctypes import c_uint64, c_float, c_char_p | |||
There was a problem hiding this comment.
Import of 'c_float' is not used.
| @@ -0,0 +1,156 @@ | |||
| import torch | |||
| import ctypes | |||
| from ctypes import c_uint64, c_float, c_double | |||
There was a problem hiding this comment.
Import of 'c_float' is not used.
1 participant
添加nvidia、metax、moore平台支持