Releases: llamara-ai/llamara-backend
Releases · llamara-ai/llamara-backend
v0.4.0
What's Changed
- Check connection & initialize embedding store on startup by @florian-h05 in #13
- Fix Qdrant default config for Docker profile by @florian-h05 in #14
- Validate embedding model config on startup by @florian-h05 in #15
- Add quarkus-info plugin to enable /q/info endpoint by @florian-h05 in #16
- Add smallrye-health plugin to enable /q/health endpoints by @florian-h05 in #17
- Add logging for creating chat model, embedding model & embedding store by @florian-h05 in #18
- Add health check to Docker image by @florian-h05 in #19
- Disable broken prompt streaming endpoint by @florian-h05 in #21
- Create payload index on Qdrant init by @florian-h05 in #22
- Dockerfile: Set OCI build-time metadata by @florian-h05 in #23
Full Changelog: v0.3...v0.4.0
v0.3
What's Changed
- Fix anonymous user cannot read public knowledge &
Users#ANY
permission cannot be added to knowledge by @florian-h05 in #2 - UserKnowledgeManagerImplTest: Add tests for modifying read-only knowledge by @florian-h05 in #4
- Move user-related functionality from
UserSessionManager
to newUserManager
by @florian-h05 in #7 - SessionManager: Refactor checkSession to throw instead of return bool by @florian-h05 in #8
KnowledgeManager#setPermission
: Don't allow settingPermission#NONE
orPermission#OWNER
by @florian-h05 in #9- Improvements to retrieving user's knowledge & Enforce authentication for knowledge management by @florian-h05 in #10
UserManager#delete
: Delete own knowledge & remove permission from shared knowledge by @florian-h05 in #11- Use Log formatting from Quarkus Log instead of String#format by @florian-h05 in #12
Full Changelog: v0.2...v0.3
v0.2
v0.1
This is the initial release of the LLAMARA backend.