Cover Image

ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Nachin, Mergen, Desai, Digant, Jia, Sicheng Stephen, Lai, Chen, Liu, Mengwei, Szwejbka, Jacob, Alvarez, Raziel, Ascani, RJ, Bort, Dave, Candales, Manuel, Caples, Andrew, Cao, Yanan, Chen, Zhengxu, Chintala, Soumith, Comer, Gregory, Islam, Tanvir, Jia, Songhao, Karuturi, Tarun, Khuu, Jack, Kukkadapu, Abhinay, Manlaibaatar, Tugsbayasgalan, Or, Andrew, Patel, Kimish, Pothapragada, Siddartha, Qiu, Lucy, Rao, Supriya, Reblitz-Richardson, Orion, Ren, Max, Roy, Scott, Shoumikhin, Anthony, Wolchok, Scott, Yang, Guang, Yi, Angela, Yuan, Martin, Zhang, Hansong, Zhang, Jack, Zhang, Jerry, Zhang, Shunting, Bilgin, C. Cagatay
Format:	Preprint
Published:	2026
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866917474229813248
author	Nachin, Mergen Desai, Digant Jia, Sicheng Stephen Lai, Chen Liu, Mengwei Szwejbka, Jacob Alvarez, Raziel Ascani, RJ Bort, Dave Candales, Manuel Caples, Andrew Cao, Yanan Chen, Zhengxu Chintala, Soumith Comer, Gregory Islam, Tanvir Jia, Songhao Karuturi, Tarun Khuu, Jack Kukkadapu, Abhinay Manlaibaatar, Tugsbayasgalan Or, Andrew Patel, Kimish Pothapragada, Siddartha Qiu, Lucy Rao, Supriya Reblitz-Richardson, Orion Ren, Max Roy, Scott Shoumikhin, Anthony Wolchok, Scott Yang, Guang Yi, Angela Yuan, Martin Zhang, Hansong Zhang, Jack Zhang, Jerry Zhang, Shunting Bilgin, C. Cagatay
author_facet	Nachin, Mergen Desai, Digant Jia, Sicheng Stephen Lai, Chen Liu, Mengwei Szwejbka, Jacob Alvarez, Raziel Ascani, RJ Bort, Dave Candales, Manuel Caples, Andrew Cao, Yanan Chen, Zhengxu Chintala, Soumith Comer, Gregory Islam, Tanvir Jia, Songhao Karuturi, Tarun Khuu, Jack Kukkadapu, Abhinay Manlaibaatar, Tugsbayasgalan Or, Andrew Patel, Kimish Pothapragada, Siddartha Qiu, Lucy Rao, Supriya Reblitz-Richardson, Orion Ren, Max Roy, Scott Shoumikhin, Anthony Wolchok, Scott Yang, Guang Yi, Angela Yuan, Martin Zhang, Hansong Zhang, Jack Zhang, Jerry Zhang, Shunting Bilgin, C. Cagatay
contents	Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model conversion or complete reimplementation outside the PyTorch ecosystem where the model was originally authored. We introduce ExecuTorch, a unified PyTorch-native deployment framework for edge AI. ExecuTorch enables seamless deployment of machine learning models across heterogeneous compute environments. It scales from embedded microcontrollers to complex system-on-chips (SoCs) with dedicated accelerators, powering devices ranging from wearables and smartphones to large compute clusters. ExecuTorch preserves PyTorch semantics while allowing customization, support for optimizations like quantization, and pluggable execution "backends". These features together enable fast experimentation, allowing researchers to validate deployment behavior entirely within PyTorch, bridging the gap between research and production.
format	Preprint
id	arxiv_https___arxiv_org_abs_2605_08195
institution	arXiv
publishDate	2026
record_format	arxiv
spellingShingle	ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device Nachin, Mergen Desai, Digant Jia, Sicheng Stephen Lai, Chen Liu, Mengwei Szwejbka, Jacob Alvarez, Raziel Ascani, RJ Bort, Dave Candales, Manuel Caples, Andrew Cao, Yanan Chen, Zhengxu Chintala, Soumith Comer, Gregory Islam, Tanvir Jia, Songhao Karuturi, Tarun Khuu, Jack Kukkadapu, Abhinay Manlaibaatar, Tugsbayasgalan Or, Andrew Patel, Kimish Pothapragada, Siddartha Qiu, Lucy Rao, Supriya Reblitz-Richardson, Orion Ren, Max Roy, Scott Shoumikhin, Anthony Wolchok, Scott Yang, Guang Yi, Angela Yuan, Martin Zhang, Hansong Zhang, Jack Zhang, Jerry Zhang, Shunting Bilgin, C. Cagatay Machine Learning Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model conversion or complete reimplementation outside the PyTorch ecosystem where the model was originally authored. We introduce ExecuTorch, a unified PyTorch-native deployment framework for edge AI. ExecuTorch enables seamless deployment of machine learning models across heterogeneous compute environments. It scales from embedded microcontrollers to complex system-on-chips (SoCs) with dedicated accelerators, powering devices ranging from wearables and smartphones to large compute clusters. ExecuTorch preserves PyTorch semantics while allowing customization, support for optimizations like quantization, and pluggable execution "backends". These features together enable fast experimentation, allowing researchers to validate deployment behavior entirely within PyTorch, bridging the gap between research and production.
title	ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device
topic	Machine Learning
url	https://arxiv.org/abs/2605.08195