Reinforcement Learning Open-Source Fest
Introducing students to open-source reinforcement learning programs and software development.
Program dates: May – August 2023
The submission portal for Reinforcement Learning (RL) Open-Source Fest 2023 is now closed.
About the program
RL Open-Source Fest is a global online program focused on introducing students to open-source reinforcement learning programs and software development while working alongside researchers, data scientists, and engineers on the Real World Reinforcement Learning team at Microsoft Research NYC. Our goal is to bring together a diverse group of students from around the world to collectively solve open-source reinforcement learning problems and advance the state-of-the-art research and development alongside the RL community while providing open-source code written and released to benefit all.
Students will work on a four-month research programming project from May-August 2023. At the end of the program, students will present each of their projects to the Microsoft Research Real World Reinforcement Learning team online.
Accepted students will receive a $10,000 USD stipend, or equivalent in local currency. Selected students can receive funding through one of the following payment options:
- Microsoft provides payment directly to a student’s academic institution, which then disperses funds according to the institution’s guidelines.
- Microsoft provides payment directly to the individual student. You will be solely responsible for any taxes related to the payments you receive from Microsoft.
Open-source projects
Vowpal Wabbit (VW) (opens in new tab) is an open-source machine learning library created by John Langford and developed by Microsoft Research with the help of many contributors. It is a fast, flexible, online, and active learning solution that empowers people to solve complex interactive machine learning problems, with a large focus on contextual bandits and reinforcement learning. It is a vehicle for both research prototyping and driving bleeding edge algorithms to production. RL OS Fest is all about open-source projects in the Vowpal Wabbit ecosystem.
View this year’s 2023 open-source projects (opens in new tab).
Eligibility
To be eligible for the program, students must be enrolled in or accepted into an accredited institution including colleges, universities, Master programs, PhD programs, and undergraduate programs.
Student responsibilities during the program
- Submit quality work: code compiles, has unit tests and documentation, and passes code review
- Regularly communicate work completed, what you intend to do next, and blockers
- Re-evaluate project tasks if you’re significantly ahead or behind schedule
- Regular check-ins with your mentor/collaborator
- Listen and respond to feedback
- Pro-active learning
What makes a successful project?
Success looks different for every project. Challenging yourself and developing skills and knowledge are the most important part. Producing some sort of deliverable item is great, but not strictly required. We all know how development and experimentation goes, unforeseen problems can come up and present new challenges and that’s all part of the process. You’ll have a mentor and support along the way.
- A successful engineering-oriented project might include pull requests merging your work, a design document, tests, and general documentation
- A successful data science-oriented project might involve pull requests, reproducible experiments, datasets, a report, and visualized results
- A successful prototyping-oriented project might include an MVP, tests, and documentation
Contact us: If you have questions about this program, please send us an email at [email protected].