Hugging Face has announced a Fellowship Program designed to support the people who’ve been quietly building the open-source machine learning ecosystem. The program is nomination-based — you can’t apply directly. Hugging Face team members or existing Fellows nominate prospects whose work has already advanced open-source ML, whether through code contributions, community building, or educational content.
The inaugural cohort shows how varied those contributions can be. María Grandury built the largest Spanish-speaking NLP community and ran a hackathon that produced 23 Spaces, 23 datasets, and 33 models pushing state-of-the-art for Spanish. Manuel Romero has contributed over 300 models to the Hub. Aritra Roy Gosthipathy added new TensorFlow architectures to Transformers. Vaibhav Srivastav led the ML4Audio working group. Bram Vanroy has been filing issues and merging pull requests since September 2019. The list goes on — each Fellow got in through a different door.
Benefits are deliberately flexible. Hugging Face offers computing resources, merchandise, and official recognition, but the company frames the real benefit as something harder to quantify: the freedom to pursue the project you’ve always wanted to do. That’s a meaningful shift from the transactional nature of most contributor programs, which tend to offer bounties for specific tasks rather than open-ended support for a person’s broader vision.
For students, there’s a separate path. The Student Ambassador Program application deadline was June 13, 2022. Hugging Face also maintains a Fellowship Doc listing strategic areas where contributions are needed, updated frequently. The company explicitly frames the program as part of its diversity and inclusion efforts, stating it won’t discriminate on any protected basis.
My read: this is Hugging Face formalizing what it was already doing informally. The nomination-only structure means it’s not a meritocracy you can game — it’s recognition of work that’s already visible to the community. If you want in, the advice is simple: start contributing. The #looking-for-contributors channel on their Discord is the on-ramp.
