Who builds NVIDIA/Megatron-LM?

18 engineers on Vamo’s public Discover lists build NVIDIA/Megatron-LM, a python repository with 18,110 stars on GitHub. The cards and the table below name each one, their role where their public card shows it, and what they did in the repo. To find every other contributor, run the search request further down or open the same search in Vamo.

Repository
NVIDIA/Megatron-LM
What it is
Ongoing research training transformer models at scale
Language
python
Stars
18,110

Builders on the Discover lists

  • May
    Jul
    Aug
    Sep
    Oct
    Nov

    Polyglot intern who ships TinyML programs for embedded devices

    Builds TinyML programs for ESP32 microcontrollers, after a program at OceanBase. Creator of TinyML-ESP32, a collection of TinyML programs for ESP32, and maintains mdbook-pdf, a backend for generating PDFs.

  • Vitaly Kurin
    @yobibyte
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    AI-first drug discovery researcher with a tooling side

    Applies ML to drug discovery at NVIDIA, after research at Isomorphic Labs. Maintains prettyplots, a tool for designing matplotlib stylefiles interactively, and builds compressor, a data compression tool.

  • oliver könig
    @ko3n1g
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Attention and transformer training MLOps engineer at NVIDIA

    Pushes changes to fast attention libraries and transformer training infrastructure, after MLOps engineering at Deutsche Post und DHL. Contributes to Dao-AILab/flash-attention, a library for exact attention, and NVIDIA/Megatron-LM, a research project for training transformer models.

  • Deepak Narayanan
    @deepakn94
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    ML systems researcher behind efficient LLM training

    Works on systems for efficient LLM training and inference, after research at Microsoft. Pushes to weld-project/weld, a runtime for data analytics apps, and contributes to mlcommons/training, a library of machine learning benchmarks.

  • Zijie Yan
    @yanring
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    LLM training engineer building MoE systems at Periodic Labs

    Builds LLM training tooling for MoE models, after leading devtech at NVIDIA. Creator of jianzhi-Offer-Leetcode, a Leetcode problem mapping library, and Megatron-MoE-ModelZoo, a model zoo for training MoE models.

  • Shiqing Fan
    @fanshiqing
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Performance architect behind large scale LLM training on NVIDIA GPUs

    Optimizes deep learning performance on NVIDIA GPUs, after development engineering at Alibaba Group. Contributes to NVIDIA/Megatron-LM, a large language model, and NVIDIA/TransformerEngine, a library for accelerating transformer models.

  • Zhenhuan Liu
    @Victarry
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    VP of Cosmos Lab, generative AI and world models

    Directs generative AI research at NVIDIA's Cosmos Lab, after principal research at Mitsubishi Electric Research Laboratories. Creator of stable-dreambooth, an implementation of Dreambooth based on Stable Diffusion, and PP-Schedule-Visualization, a visualization tool for pipeline parallelism.

  • Keval Morabia
    @kevalmorabia97
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Staff deep learning engineer pushing NVIDIA's speech AI and model optimization

    Ships generative AI speech and model optimization tooling at NVIDIA, after content creation at Instagram. Commits to NVIDIA-NeMo/Speech, a framework for generative AI, and to NVIDIA/Model-Optimizer, a library for model optimization.

  • Xuanteng Huang
    @xuantengh
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Heterogeneous compute engineer on LLM training at NVIDIA

    Architects GPU compute systems for LLM training infrastructure at NVIDIA, after recommendation systems at Weixin/WeChat. Contributes to python/cpython, the Python programming language.

  • Xin Yao
    @yaox12
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Software engineer scaling transformer training for NVIDIA GPUs

    Delivers transformer training and graph deep learning tooling at NVIDIA. Contributes to dmlc/dgl, a library for deep learning on graphs, and NVIDIA/Megatron-LM, a research project for training transformer models at scale.

  • Yan Bai
    @ISEEKYAN
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    RL post-training engineer shipping frameworks at NVIDIA

    Delivers reinforcement learning post-training frameworks at NVIDIA, after software engineering at ByteDance. Contributes to verl-project/verl, a framework for post-training reinforcement learning models.

  • John St. John
    @jstjohn
    May
    Jul
    Aug
    Sep
    Oct
    Nov

    Principal AI scientist bridging drug discovery and speech models at NVIDIA

    Applies AI to drug discovery at NVIDIA, after applied science at Amazon. Contributes to NVIDIA-NeMo/Speech, a framework for building speech and language models, and NVIDIA-BioNeMo/bionemo-recipes, a recipe repository for AI models in drug discovery.

NameGitHubRoleWork in this repo
ℍ𝕠𝕝𝕝𝕠𝕨 𝕄𝕒𝕟HollowMan6Deep Learning Algorithm Engineering InternContributes to NVIDIA/Megatron-LM.
Vitaly KurinyobibyteSenior Research ScientistContributes to NVIDIA/Megatron-LM.
oliver königko3n1gSenior MLOps EngineerContributes to NVIDIA/Megatron-LM.
Deepak Narayanandeepakn94Senior Applied Deep Learning Research ScientistContributes to NVIDIA/Megatron-LM.
Zijie YanyanringMember of Technical StaffContributes to NVIDIA/Megatron-LM.
Shiqing FanfanshiqingSenior Performance ArchitectContributes to NVIDIA/Megatron-LM.
Zhenhuan LiuVictarryVice President of Cosmos LabContributes to NVIDIA/Megatron-LM.
Keval Morabiakevalmorabia97Staff Deep Learning EngineerContributes to NVIDIA/Megatron-LM.
Xuanteng HuangxuantenghCompute Arch Intern, LLM trainingContributes to NVIDIA/Megatron-LM.
Xin Yaoyaox12Senior Software EngineerContributes to NVIDIA/Megatron-LM.
Yan BaiISEEKYAN工程师Contributes to NVIDIA/Megatron-LM.
John St. JohnjstjohnPrincipal Drug Discovery AI Applied ScientistContributes to NVIDIA/Megatron-LM.
Anil ThomasanlthmsPrincipal Research ScientistContributes to NVIDIA/Megatron-LM.
Jared CasperjaredcasperSenior Deep Learning Research ScientistContributes to NVIDIA/Megatron-LM.
Kirthi Shankar SivamaniksivamanSenior Deep Learning EngineerContributes to NVIDIA/Megatron-LM.
Eric W. Trameleric-tramelMember of Technical StaffContributes to NVIDIA/Megatron-LM.
Boxiang WangBoxiangWSenior Research EngineerContributes to NVIDIA/Megatron-LM.
Santosh BhavanisbhavaniSr. Software Product Manager, Deep LearningContributes to NVIDIA/Megatron-LM.

The lists show a few people. This request to the developer search API returns everyone Vamo has seen contribute to the repository, ranked by devRank. It is the same search the app runs when you search for NVIDIA/Megatron-LM contributors in Vamo. The API docs cover every filter, and the repository index lists the other repos with builders on the lists.

POST /v2/developers/search
{  "repos": {    "names": [      "NVIDIA/Megatron-LM"    ],    "role": "contributor"  },  "filters": [    "And",    [      [        "is_bot",        "Eq",        false      ],      [        "is_org_account",        "Eq",        false      ]    ]  ],  "sort": "devrank"}
“Contributors to NVIDIA/Megatron-LM” as one request. The response lists each person and what matched.

Hire the people who build NVIDIA/Megatron-LM.