Repository files navigation

ExecuTorch: A powerful on-device AI Framework

---

ExecuTorch is an end-to-end solution for on-device inference and training. It powers much of Meta's on-device AI experiences across Facebook, Instagram, Meta Quest, Ray-Ban Meta Smart Glasses, WhatsApp, and more.

It supports a wide range of models including LLMs (Large Language Models), CV (Computer Vision), ASR (Automatic Speech Recognition), and TTS (Text to Speech).

Platform Support:

• Operating Systems:

  • iOS
  • MacOS (ARM64)
  • Android
  • Linux
  • Microcontrollers

• Hardware Acceleration:

  • Apple
  • Arm
  • Cadence
  • MediaTek
  • OpenVINO
  • Qualcomm
  • Vulkan
  • XNNPACK

Key value propositions of ExecuTorch are:

• Portability: Compatibility with a wide variety of computing platforms,from high-end mobile phones to highly constrained embedded systems andmicrocontrollers.
• Productivity: Enabling developers to use the same toolchains and DeveloperTools from PyTorch model authoring and conversion, to debugging and deploymentto a wide variety of platforms.
• Performance: Providing end users with a seamless and high-performanceexperience due to a lightweight runtime and utilizing full hardwarecapabilities such as CPUs, NPUs, and DSPs.

To get started you can:

• Visit theStep by Step Tutorial to get things running locally and deploy a model to a device
• Use thisColab Notebook to start playing around right away
• Jump straight into LLM use cases by following specific instructions for popular open-source models such asLlama,Qwen 3,Phi-4-mini, andLlava

We welcome any feedback, suggestions, and bug reports from the community to helpus improve our technology. Check out theDiscussion Board or chat real time with us onDiscord

We welcome contributions. To get started review theguidelines and chat with us onDiscord

Please refer to theCodebase structure section of theContributing Guidelines for more details.

ExecuTorch is BSD licensed, as found in the LICENSE file.