Why

After investigating the benefits of fine-tuning an LLM, we've identified a lack of Ruby code in the training data of most foundational open source models. This lack of Ruby in these models greatly limits the quality of the code they output. To solve this, we can take advantage of fine-tuning, and significantly improve the presence and joy of Ruby in AI.

Project Goals

1. Fine-Tuning an LLM for Ruby: This model will be tailored to reflect the Ruby community’s preferences for high quality, elegant design, and preferred syntax.
2. Open Source and Community-Driven: Training will leverage the wealth of open-source Ruby projects, comprehensive documentation, how-to guides, and other valuable community-contributed content. All content used for training will obtain proper consent from the contents’ owner(s), and fully documented.
3. Building in Public: Transparency is key. Every step of the development process will be shared, encouraging collaboration and feedback from the Ruby community.