Apache TVM Documentation

Welcome to the documentation for Apache TVM, a deep learning compiler that enables access to high-performance machine learning anywhere for everyone. TVM’s diverse community of hardware vendors, compiler engineers and ML researchers work together to build a unified, programmable software stack, that enriches the entire ML technology ecosystem and make it accessible to the wider ML community. TVM empowers users to leverage community-driven ML-based optimizations to push the limits and amplify the reach of their research and development, which in turn raises the collective performance of all ML, while driving its costs down.

Getting Started

• Overview
• Installing TVM
• Quick Start
• IRModule

How To

• End-to-End Optimize Model
• Customize Optimization
• Optimize Large Language Model
• Cross Compilation and RPC
• Development Guides

Deep Dive

• Design and Architecture
  • Overall Flow
  • tvm/support
  • tvm/runtime
  • tvm/node
  • tvm/ir
  • tvm/target
  • tvm/relax
  • tvm/tir
  • tvm/arith
  • tvm/te and tvm/topi
  • tvm/meta_schedule
  • tvm/dlight
• TensorIR
  • Tensor Program Abstraction
  • Understand TensorIR Abstraction
  • TensorIR Creation
  • Transformation
• Relax
  • Graph Abstraction for ML Models
  • Understand Relax Abstraction
  • Relax Creation
  • Transformation

API Reference

• Python API
• Other APIs

Legacy

• User Tutorial
• How To Guides
• Developer Tutorial
• Developer How-To Guide
• Language Reference
• microTVM: TVM on bare-metal
• VTA: Versatile Tensor Accelerator

About

• Contributor Guide
• Publications
• Security Guide

Index

• Index