PeopleofSuperintelligence

Builder

Lu Fang

As of October 2026, Lu Fang is a research scientist at Meta who has worked on open-source AI software, including PyTorch, ONNX and Caffe2. By 2018 he held a PhD from UC Irvine and worked on Facebook's AI Platform team, where he helped develop Caffe2 and PyTorch. He also led the effort on ONNX, an open format that bridges different AI frameworks. The paper describing PyTorch, which he co-authored, appeared in the NeurIPS 2019 proceedings. In November 2025 Soumith Chintala named him among the people who became centrally important to PyTorch later.

Now at
Meta
Works
197

Publication numbers from OpenAlex, as of 15 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. By 2018

    Research scientist at Facebook AI

    On Facebook's AI Platform team, he works as a research scientist and helps develop Caffe2 and PyTorch.

  2. By 2018

    The ONNX exchange format

    The ONNX project, an open format that bridges different AI frameworks, is one he leads.

  3. By 2018

    A PhD from UC Irvine

    He holds a PhD from UC Irvine, plus master's and bachelor's degrees from Peking University.

  4. November 2018

    Talk on PyTorch 1.0 and ONNX

    At UCLA, he speaks on moving research into production with PyTorch 1.0 and ONNX.

  5. 2019

    Co-author of the PyTorch paper

    The paper that describes PyTorch, which he co-authors, appears in the NeurIPS 2019 proceedings.

  6. November 2025

    Named a later PyTorch contributor

    Soumith Chintala names him among the people who became centrally important to PyTorch later.

  7. As of October 2026

    Research scientist at Meta

    At Meta, he is a research scientist and has worked on open-source projects including vLLM, PyTorch, ONNX and Caffe2.

Sources and credits

Photo: We have no published photo of Lu Fang yet. Send us a free-licensed photo.

  1. CS 201: Accelerating Research to Production with PyTorch 1.0 and ONNX, LU FANG, Facebook AI | CS cs.ucla.edu. Checked 15 October 2026.
  2. PyTorch: An Imperative Style, High-Performance Deep Learning Library proceedings.neurips.cc. Checked 15 October 2026.
  3. [1912.01703] PyTorch: An Imperative Style, High-Performance Deep Learning Library arxiv.org. Checked 15 October 2026.
  4. Leaving Meta and PyTorch soumith.ch. Checked 15 October 2026.
  5. Lu Fang's Speaker Profile @ Sessionize sessionize.com. Checked 15 October 2026.