Sign In
Get Clay Free →

Suggestions

    Jason Ansel

    PyTorch Compilers

    Jason Ansel is a Principal Engineer (L8) at Meta, where he focuses on advancing machine learning technologies, particularly through his work with PyTorch. He is recognized as a meticulous engineer who has made significant contributions to optimizing code sequences in various projects.

    Professional Background

    • Current Position: Principal Research Scientist at Meta since April 2020. In this role, he leads the development of PyTorch compilers and has initiated key projects like TorchDynamo and TorchInductor, which enhance the performance of PyTorch by enabling faster training and inference through dynamic Python bytecode transformation.126
    • Previous Experience:
      • GoDaddy: Held several roles from Director of Engineering to Distinguished Engineer between 2013 and 2020, where he contributed to building a deep learning platform and developed tools for domain name appraisal using neural networks.13
      • Locu, Inc.: Served as Director of Machine Learning from 2011 to 2013 before its acquisition by GoDaddy.1
      • Internships: Ansel has also interned at notable companies including Google and AMD, gaining diverse experience in engineering and technology.1

    Education

    Jason Ansel earned his Ph.D. from the Massachusetts Institute of Technology (MIT) in 2014, focusing on the intersection of machine learning, compilers, and programming languages. His dissertation involved developing the OpenTuner project and the PetaBricks programming language, aimed at automating program optimization.45

    Contributions to PyTorch

    Ansel's work on PyTorch has been pivotal:

    • TorchDynamo: A Just-In-Time (JIT) compiler that optimizes unmodified PyTorch programs for better performance.
    • TorchInductor: A new compiler backend that translates PyTorch operations into efficient code for both GPUs and CPUs.26

    His efforts have resulted in significant performance improvements for machine learning tasks, making PyTorch a more powerful tool for researchers and developers alike.

    Online Presence

    For more detailed insights into his work and projects, Jason Ansel can be found on LinkedIn and GitHub where he shares his repositories and contributions to open-source software.

    Highlights

    Feb 16 · acronymattic.com
    PAV - Pennine Aim Vct | AcronymAttic
    Jan 17 · data.sfgov.org
    https://data.sfgov.org/api/views/wr8i-kpa5/rows.cs...
    Nov 13 · youtube.com
    Research Seminar: Jason Ansel (Meta AI) - YouTube
    Research Seminar: Jason Ansel (Meta AI) - YouTube
    Apr 30 · mlsys-ucsd.org
    PyTorch 2: Faster Machine Learning Through Dynamic Python ...
    Spatially organized multicellular immune hubs in human colorectal ...
    Jan 1 · github.com
    Jason Ansel jansel - GitHub
    Jul 26 · jasonansel.com
    Jason Ansel
    Jason Ansel

    Related Questions

    What are Jason Ansel's main contributions to PyTorch 2?
    How did Jason Ansel's work at GoDaddy influence his career at Meta?
    What is the significance of TorchDynamo and TorchInductor in PyTorch 2?
    Can you provide more details about Jason Ansel's Ph.D. research at MIT CSAIL?
    What are some of the notable projects Jason Ansel has led at Meta?
    Jason Ansel
    Jason Ansel, photo 1
    Jason Ansel, photo 2
    Add to my network

    Location

    San Francisco Bay Area