Skip to content
forked from apache/tvm

Open deep learning compiler stack for cpu, gpu and specialized accelerators

License

Notifications You must be signed in to change notification settings

giuseros/incubator-tvm

 
 

Repository files navigation

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.

About

Open deep learning compiler stack for cpu, gpu and specialized accelerators

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 50.5%
  • C++ 41.0%
  • Scala 1.4%
  • Rust 1.3%
  • Java 1.2%
  • Go 0.8%
  • Other 3.8%