apache/tvm is a machine learning compiler framework that tackles the complex, cross-level compilation pipeline needed to optimize and deploy models. It uses Python-first development to bring models into minimum deployable modules.
Project overview
apache/tvm addresses the complex challenge of cross-level model compilation with a cross-level representation that can jointly optimize computational graphs, tensor programs, and libraries, and serves as foundation infrastructure for building domain-specific compilers.
Project type
Model Runtime · Infrastructure
Deployment
Refer to project documentation
License
Apache-2.0
Best for
AI engineers and researchers who need to optimize models and are comfortable writing code to customize compiler pipelines.
Developers building Python-first vertical compilers for specific machine learning domains.
Key capabilities
Enables quick customization of machine learning compiler pipelines using Python.
Brings models into minimum deployable modules.
Provides a cross-level representation that can jointly optimize computational graphs, tensor programs, and libraries.
Serves as foundation infra for building Python-first vertical compilers for domains, such as LLMs.
Getting started
Setup difficulty is moderate because it requires building or installing a compiler framework and writing code to use the library. To get started, check out the TVM Documentation site for installation instructions.
Evidence and sources
GitHub project description: Open Machine Learning Compiler Framework
README: Apache TVM is an open machine learning compilation framework, following the following principles: - Python-first development that enables quick customization of machine learning c…
README: The project's current design goal is to make the ML compiler accessible by enabling most transformations to be customizable in Python and bringing a cross-level representation tha…
README: Check out the TVM Documentation site for installation instructions, tutorials, examples, and more.
README: TVM started as a research project for deep learning compilation.