Machine Learning Engineer Internship, Hugging Face
TRL
Community-driven platform for ML models
Job no longer available
Community-driven platform for ML models
201-500 employees
Job no longer available
201-500 employees
Company mission
Hugging Face is on a mission to democratize good machine learning, one commit at a time.
Role
Who you are
- Machine Learning: Fine-tuning large language models (LLMs) or vision-language models (VLMs), and optimisation techniques
- Software Development: Proficiency in Python, PyTorch, and frameworks like Hugging Face Transformers, with experience in distributed training and GPU acceleration
- Open-Source: Familiarity with Git/GitHub workflows, community engagement, documentation, and collaborative development
- Research and Experimentation: Exposure to cutting-edge ML research, benchmarking, and testing fine-tuning methods
- Tooling and Maintenance: Building tools to streamline workflows, ensuring software stability, backward compatibility, versioning, and delivering reliable releases
- Communication and Outreach: Writing blog posts, tutorials, and sharing updates on platforms like LinkedIn to engage with the community and make complex concepts accessible to a wider audience
- You’re passionate about open-source innovation and making advanced ML tools accessible globally
- You value continuity in software development, ensuring users have a dependable and evolving library to rely on
What the job involves
- In the past year the focus of building LLMs has gradually shifted from pretraining to post-training
- This means spending more and more time on figuring out how to get models follow instructions reliably, use tools and generally align with certain values
- With over 10k Github stars and close to 1M monthly installs the has become one of the go-to libraries for post-training
- It scales flexibly from a single GPU to large clusters of GPUs using PEFT and ZeRO and offers a wide range of trainers for the latest post training techniques such as PPO or DPO and many more
- In addition it includes a user friendly CLI that allows training models with a single command
- During this internship, you will collaborate with the research team to integrate cutting-edge methods into the library, maintain a clean and scalable codebase, and ensure its usability through thoughtful documentation
- You’ll actively engage with the TRL community by responding to issues, gathering feedback, and fostering collaboration through thoughtful discussions and support, ensuring the library continues to meet developers' needs
- Your contributions will directly influence thousands of developers globally, advancing the adoption of state-of-the-art post-training techniques and laying the groundwork for the next generation of customizable, instruction-following LLMs
Application process
- 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
Our take
Large language models, which recognize and generate text, are already being used to auto-complete sentences, power chat bots, moderate content, summarize news articles, and translate text online. However, many of these advances are happening behind closed doors, stifling innovation and raising ethicacy concerns.
To combat this, Hugging Face is building the "GitHub of machine learning". It’s a community-driven platform with a ton of repositories, where developers can create, discover and collaborate on ML models, datasets and ML apps. This work led to the creation of BLOOM in 2022, an open-access LLM with similar architecture to OpenAI’s GPT, but that is multilingual by design.
The company also launched the Hugging Chat Assistant, an open-source tool to create personalised chat assistants, in direct competition to OpenAI's paid service ChatGPT Plus. By continually creating value based on a global community, not just for Microsoft or Google, HuggingFace should be able to carve out its own path for success.
Freddie
Company Specialist at Welcome to the Jungle
Insights
36% employee growth in 12 months
Company
Funding (last 2 of 7 rounds)
Aug 2023
$235m
SERIES D
May 2022
$100m
SERIES C
Total funding: $395.8m
Company benefits
- Flexible Work Environment
- Health Insurance
- Unlimited PTO
- Equity
- Growth, Training, & Conferences
- Generous Parental Leave
Company HQ
Dumbo, NY
Leadership
Clem Delangue
(CEO)Created Moodstocks, a startup building machine learning for computer vision acquired by Google. Also built Mention where they were head of Product and CMO until the company was acquired by MyNewsDesk.
Julien Chaumond
(CTO)Former advisor for the Deputy Minister for Digital Affairs in Paris, and Senior engineer for Stupefix app that was acquired by GoPro. Also co-founded and was CTO of Glose.
Thomas Wolf
(CSO)After graduating from Ecole Polytechnique, they worked on laser-plasma interactions at the BELLA Center of the Lawrence Berkeley National Laboratory. Also worked as a European Patent Attorney for 5 years.
Share this job
View 5 more jobs at Hugging Face