The Torch-MLIR project aims to provide first class support from the PyTorch ecosystem to the MLIR ecosystem.
 
 
 
 
 
 
Go to file
武家伟 636f5acb10
[MHLO] Init MHLO reduce-like op conversion (#1133)
* [MHLO] init reduce-like op conversion from Torch to MHLO
Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
2022-08-03 10:47:52 +08:00
.github Resolve remaining LTC CI failures (#1110) 2022-07-30 09:40:02 -04:00
build_tools Resolve remaining LTC CI failures (#1110) 2022-07-30 09:40:02 -04:00
docs Reference Lazy Backend (#1045) 2022-07-30 09:40:02 -04:00
e2e_testing/torchscript Add embedding Bag e2e case in xfail set (#1130) 2022-08-01 18:23:45 -04:00
examples Reference Lazy Backend (#1045) 2022-07-30 09:40:02 -04:00
externals llvm: update tag to 02b3a35 (#1124) 2022-07-31 15:26:24 -07:00
include Add support for `dim=None` to `AtenMeanDimOp` (#1129) 2022-08-02 16:08:06 +00:00
lib [MHLO] Init MHLO reduce-like op conversion (#1133) 2022-08-03 10:47:52 +08:00
python Add support for `dim=None` to `AtenMeanDimOp` (#1129) 2022-08-02 16:08:06 +00:00
test [MHLO] Init MHLO reduce-like op conversion (#1133) 2022-08-03 10:47:52 +08:00
tools build: improve robustness of cmake and shell scripts (#1018) 2022-07-06 14:39:30 -07:00
utils/bazel Fix mhlo bazel rule name (#1136) 2022-08-02 12:28:53 -07:00
.clang-format Add stub numpy dialect. 2020-04-26 17:20:58 -07:00
.gitignore Reference Lazy Backend (#1045) 2022-07-30 09:40:02 -04:00
.gitmodules [MHLO] Init MHLO integration. (#1083) 2022-07-20 16:18:16 -07:00
.style.yapf Change preferred style to be PEP8 2022-04-20 14:38:19 -07:00
CMakeLists.txt Resolve remaining LTC CI failures (#1110) 2022-07-30 09:40:02 -04:00
LICENSE Dual license the torch-mlir project. 2021-10-01 10:46:08 -07:00
README.md LTC Documentation (#1021) 2022-07-30 09:40:02 -04:00
Torch-MLIR.png Update diagram for TOSA backend. 2022-04-01 22:46:25 +00:00
development.md Update development.md with source builds (#1105) 2022-07-25 10:24:45 -07:00
pyproject.toml Minor buildsystem fixes (#778) 2022-04-21 15:53:00 -07:00
requirements.txt Revert requirements.txt (#930) 2022-06-10 15:23:12 -07:00
setup.py Disable LTC on Release builds (#1125) 2022-07-31 20:58:29 -07:00

README.md

The Torch-MLIR Project

The Torch-MLIR project aims to provide first class compiler support from the PyTorch ecosystem to the MLIR ecosystem.

This project is participating in the LLVM Incubator process: as such, it is not part of any official LLVM release. While incubation status is not necessarily a reflection of the completeness or stability of the code, it does indicate that the project is not yet endorsed as a component of LLVM.

PyTorch An open source machine learning framework that accelerates the path from research prototyping to production deployment.

MLIR The MLIR project is a novel approach to building reusable and extensible compiler infrastructure. MLIR aims to address software fragmentation, improve compilation for heterogeneous hardware, significantly reduce the cost of building domain specific compilers, and aid in connecting existing compilers together.

Torch-MLIR Multiple Vendors use MLIR as the middle layer, mapping from platform frameworks like PyTorch, JAX, and TensorFlow into MLIR and then progressively lowering down to their target hardware. We have seen half a dozen custom lowerings from PyTorch to MLIR. Having canonical lowerings from the PyTorch ecosystem to the MLIR ecosystem would provide much needed relief to hardware vendors to focus on their unique value rather than implementing yet another PyTorch frontend for MLIR. The goal is to be similar to current hardware vendors adding LLVM target support instead of each one also implementing Clang / a C++ frontend.

Release Build

All the roads from PyTorch to Torch MLIR Dialect

We have few paths to lower down to the Torch MLIR Dialect.

Torch Lowering Architectures

  • TorchScript This is the most tested path down to Torch MLIR Dialect, and the PyTorch ecosystem is converging on using TorchScript IR as a lingua franca.
  • LazyTensorCore Read more details here.

Project Communication

  • #torch-mlir channel on the LLVM Discord - this is the most active communication channel
  • Github issues here
  • torch-mlir section of LLVM Discourse
  • Weekly meetings on Mondays 9AM PST. See here for more information.
  • Weekly op office hours on Thursdays 8:30-9:30AM PST. See here for more information.

Install torch-mlir snapshot

This installs a pre-built snapshot of torch-mlir for Python 3.7/3.8/3.9/3.10 on Linux and macOS.

python -m venv mlir_venv
source mlir_venv/bin/activate
# Some older pip installs may not be able to handle the recent PyTorch deps
python -m pip install --upgrade pip
pip install --pre torch-mlir torchvision -f https://github.com/llvm/torch-mlir/releases --extra-index-url https://download.pytorch.org/whl/nightly/cpu
# This will install the corresponding torch and torchvision nightlies

Demos

TorchScript ResNet18

Standalone script to Convert a PyTorch ResNet18 model to MLIR and run it on the CPU Backend:

# Get the latest example if you haven't checked out the code
wget https://raw.githubusercontent.com/llvm/torch-mlir/main/examples/torchscript_resnet18.py

# Run ResNet18 as a standalone script.
python examples/torchscript_resnet18.py

load image from https://upload.wikimedia.org/wikipedia/commons/2/26/YellowLabradorLooking_new.jpg
Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /home/mlir/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
100.0%
PyTorch prediction
[('Labrador retriever', 70.66319274902344), ('golden retriever', 4.956596374511719), ('Chesapeake Bay retriever', 4.195662975311279)]
torch-mlir prediction
[('Labrador retriever', 70.66320037841797), ('golden retriever', 4.956601619720459), ('Chesapeake Bay retriever', 4.195651531219482)]

Lazy Tensor Core

View examples here.

Eager Mode

Eager mode with TorchMLIR is a very experimental eager mode backend for PyTorch through the torch-mlir framework. Effectively, this mode works by compiling operator by operator as the NN is eagerly executed by PyTorch. This mode includes a fallback to conventional PyTorch if anything in the torch-mlir compilation process fails (e.g., unsupported operator). A simple example can be found at eager_mode.py. A ResNet18 example can be found at eager_mode_resnet18.py.

Repository Layout

The project follows the conventions of typical MLIR-based projects:

  • include/torch-mlir, lib structure for C++ MLIR compiler dialects/passes.
  • test for holding test code.
  • tools for torch-mlir-opt and such.
  • python top level directory for Python code

Developers

If you would like to develop and build torch-mlir from source please look at Development Notes