Llama 3.1 and Gemma 2 Support
Jan finally supports Meta's Llama 3.1 and Google's Gemma 2. Thanks for the patience folks!
We've been working on stability issues over the last few weeks. Jan is now more stable.
Additional Notes
- Upgraded the inference engine for better performance
- Model settings now persist across new threads
- Fixed the GPU memory utilization bar
- Some UX and copy improvements
For more details, see the GitHub release notes (opens in a new tab).