Skip to content
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Python C++ Scala Rust Java Go Other
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github
3rdparty
apps
cmake
conda
docker
docs
golang
include/tvm
jvm
licenses
nnvm [LINT] Remove unnecessary copyright message for files with ASF header ( Nov 24, 2019
python
rust
src
tests
topi
tutorials
vta
web
.clang-format
.gitignore
.gitmodules
CMakeLists.txt
CONTRIBUTORS.md
DISCLAIMER
Jenkinsfile
KEYS
LICENSE
Makefile
NEWS.md
NOTICE
README.md
version.py

README.md

Open Deep Learning Compiler Stack

Documentation | Contributors | Community | Release Notes

Build Status Azure Pipeline

Apache TVM (incubating) is a compiler stack for deep learning systems. It is designed to close the gap between the productivity-focused deep learning frameworks, and the performance- and efficiency-focused hardware backends. TVM works with deep learning frameworks to provide end to end compilation to different backends.

License

© Contributors Licensed under an Apache-2.0 license.

Contribute to TVM

TVM adopts apache committer model, we aim to create an open source project that is maintained and owned by the community. Checkout the Contributor Guide

Acknowledgement

We learned a lot from the following projects when building TVM.

  • Halide: TVM uses HalideIR as data structure for arithmetic simplification and low level lowering. We also learned and adapted some part of lowering pipeline from Halide.
  • Loopy: use of integer set analysis and its loop transformation primitives.
  • Theano: the design inspiration of symbolic scan operator for recurrence.
You can’t perform that action at this time.