Data Science Working Group
This repository will be used to track all the work that comes out of the Red Hat Data Science Working Group.
Purpose
The purpose of this working group is to bring together a group of contributors interested in evolving the different areas of data science at Red Hat. It aims to foster collaboration, share knowledge, and drive innovation in the field of data science. The group is open to all contributors who have interest in this work. Some key objectives would be:
- Knowledge Sharing and Collaboration - Organize regular meetings to discuss current topics, challenges, and innovations in data science. Encourage collaboration among members from different domains to promote the exchange of ideas and expertise.
- Research and Development - Highlight and explore potential research areas within data science that can contribute to advancements in the field. Facilitate collaborative research projects among members, providing a platform for collective problem-solving and innovation
- Community Building - Build a supportive community where members can seek advice, share experiences, and provide mutual support.
Areas of Interest
Initial Focus Areas
To begin with, the Data Science working group will concentrate on the following key areas. These areas have been identified based on their current relevance and potential for impact at Red Hat:
- InstructLab - The InstructLab project has scope for improvement in various aspects such as model evaluation, model benchmarking, model fine tuning and other data science areas that can be enhanced. We would like to look into some of these areas and conduct experiments, POCs and prototypes that address some of the gaps in the current InstructLab implementations and share our findings to the relevant upstream/downstream teams.
Evolution of Areas of Interest
The working group will adjust its focus areas to stay aligned with emerging trends and technologies, maintaining the flexibility to pivot according to Red Hat's priorities. The group will regularly review and update its areas of interest based on member feedback and outcomes from joint collaborative projects across the company.
Group Membership
There is no formal membership of this group. Those interested in contributing just need to show up and participate in the work of the group.
Community
- Meeting - Weekly on Monday, 12:30pm EST
- Slack - We have a slack channel for this group which you can join: #datascience-wg
- Mailing List - TBD