In this update:
- Added the SambaNova provider. It offers extremely fast Llama-3.1 405B, 70B and 8B models (up to 1000 tokens/sec during testing).
- Enhancements for Local API configuration.
What's Changed
- Add SambaNova provider by @XInTheDark in #86
Full Changelog: v2.7.1...v2.8