Open End to End AI Compiler Stack

GitHub license Build Status

TVM stack is a unified optimization stack that will close the gap between the productivity-focused deep learning frameworks, and the performance- or efficiency-oriented hardware backends. The project contains the following components:

  • TVM Tensor level compiler Stack for Deep Learning Systems
  • NNVM Graph compiler stack for Deep Learning Systems
  • NNVM compiler: open compiler for AI Frameworks (shares same repo with NNVM)

Checkout our techreport TVM: End-to-End Optimization Stack for Deep Learning