Hugging Face Fellowship Program Announcement

Hugging Face has introduced the Fellowship Program, a network designed to empower exceptional contributors to the open-source machine learning (ML) ecosystem. The program aims to scale the impact of key contributors and inspire others to participate in the democratization of machine learning.

Program Objectives and Structure

The Hugging Face Fellowship is designed to support grassroots contributions to machine learning. Rather than following a rigid structure, the program provides tailored support based on the individual needs of each Fellow to help them execute projects they have always wanted to pursue.

Fellow Benefits

Support for Fellows is customized to their specific interests and projects. Examples of provided benefits include:

  • Computing and resources: Access to necessary technical infrastructure.
  • Merchandise and assets: Physical goods and tools.
  • Official recognition: Formal acknowledgement from Hugging Face.

Nomination and Selection Criteria

Admission to the Fellowship is ongoing and based on nominations from current Fellows or members of the Hugging Face team. The primary criterion for selection is a demonstrated contribution to the democratization of open-source machine learning.

Examples of Fellow Contributions

The program recognizes a diverse range of contributions, including:

  • Community Building: María Grandury created the largest Spanish-speaking NLP community and organized a hackathon resulting in 23 Spaces, 23 datasets, and 33 models.
  • Model Development: Manuel Romero contributed over 300 models to the Hugging Face Hub, including several SOTA models in Spanish.
  • Technical Contributions: Aritra Roy Gosthipathy contributed new TensorFlow architectures to the Transformers library and improved Keras tooling.
  • Advocacy and Education: Vaibhav Srivastav led the ML4Audio working group and conducted paper discussion sessions.
  • Long-term Maintenance: Bram Vanroy has contributed issues and pull requests to the Transformers library since September 2019.
  • Project Implementation: Christopher Akiki contributed to Big Science, sprints, and workshops, creating projects such as TF-coder.
  • Space Development: Ceyda DZnarel developed successful Hugging Face Spaces, including the ButterflyGAN Space.

How to Contribute to Open-Source ML

For those not yet in the Fellowship, Hugging Face suggests several pathways to contribute to the open-source ML ecosystem:

  • Model Sharing: Uploading Computer Vision, Reinforcement Learning, or other ML domain models to the Hub.
  • Tutorials and Documentation: Creating projects using libraries like fastai, Keras, or Stable-Baselines 3, or translating the Hugging Face Course and Transformers documentation.
  • Community Organization: Organizing local sprints to promote ML in specific languages or niches.
  • Strategic Projects: Contributing to specific projects listed in the official Fellowship documentation.

Program FAQ and Eligibility

  • Employment Status: Fellows are not employees of Hugging Face.
  • Student Eligibility: Students are encouraged to apply to the Student Ambassador Program (which had an application deadline of June 13, 2022).
  • Timeline: There is no fixed deadline for admission; it is an ongoing process based on nomination.
  • Diversity and Inclusion: The program explicitly prohibits discrimination based on race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.

Sources