Lead: This working group is led by Jeremy Fowers, whose handle is @jeremyfowers on GitHub and @jfowers_amd on Discord.
Background: Lemonade Omni Models combine multiple models and backends together to present omni-modal capabilities (e.g., chat and image editing in the same model) that would otherwise not be readily available in local AI.
Why: Omni-modality enables natural interactions between end-users and local AI systems, and we believe it is essential for mass adoption of local AI.
Goal: Improve Omni Models to become a staple of local AI apps and use. The condition for this is not yet apparent, so we will start by iterating on some obvious next steps.
Please see the general contribution guidelines, then contact @jfowers_amd on Discord before getting started to discuss the roadmap.
Roadmap items may be high-level objectives that may span multiple issues and PRs. Specific issues and PRs can be found in the Omni Models GitHub Milestone.