About the Role

As an Open-Source Machine Learning Engineer, you'll work to improve the open-source machine learning ecosystem. You'll mainly work on existing open-source libraries such as Transformers, Datasets, Pytorch and vLLM, and you'll interact with users and contributors across the broad open-source ML ecosystem. We'll brainstorm with you to put you in a position to do the work that interests you and that is impactful.

You'll help foster one of the most active machine learning communities, helping users contribute to and use the tools you build. You'll work with researchers, ML practitioners, and data scientists every day through GitHub, our forums, and Slack.

What you'll need

Strong Python skills, with experience writing clean, well-tested, maintainable library code
Deep hands-on experience with a modern deep-learning framework, especially PyTorch (JAX or TensorFlow a plus)
Practical experience with the Hugging Face open-source stack (Transformers, Datasets, Accelerate) or comparable ML libraries
A public track record of open-source contributions, for example merged pull requests to ML or data libraries, that we can review on GitHub
Solid understanding of modern machine learning and deep learning, including transformer architectures
Experience collaborating with a technical community in the open (GitHub issues and reviews, forums, Slack or Discord)
Fluent written English for asynchronous collaboration across a distributed, global community

Nice to have

Experience maintaining an open-source project
Prior contributions to Transformers, Datasets, Accelerate, or similar libraries
Familiarity with distributed training, inference optimization, or GPU/accelerator performance work
Experience training or fine-tuning models at scale

Requirements

Please provide a cover letter mentioning why you would like to work in open-source at Hugging Face. We encourage you to mention your skills, potential expertise, and topics on which you would like to work.