v1.8.0

What's new

• Added support for GPT-4 Omni mini
• Added support for Anthropic models
• Added Gemma2 9B
• Separated all the providers in different files

Fixes

• Fixes invalid provider issue
• Fixes infinity loading & web socket connection errors
• Fixes build errors; reverted base image back to node:slim

v1.7.1

What's changed?

• Fixed Docker image build errors
• Added chat not found screen
• Added chat deletion functionality

v1.7.0

What's Changed

• Implement WS error handling
• Fix minor UI bugs
• Add library feature
• Add permanent data store
• Removed unimplemented discover page

Full Changelog: v1.6.1...v1.7.0

v1.6.1

Changelogs

• Fix build error due to invalid parameters

v1.6.0

What's Changed

• Add restart policy in docker-compose.yaml
• Fix: add extra_hosts to docker-compose.yaml to allow connection to Ollama
• Adapt light mode

Full Changelog: v1.5.0...v1.6.0

v1.5.0

v1.5.0 Release Notes

What's Changed

• Bug Fixes

  • Fixed "list map element must specify a unique key" issue.
  • Corrected the baseURL storage key in the settings dialog.
  • Resolved suspense boundary errors in the chat window.

• New Features

  • Introduced support for suggestions.
  • Added ListLineOutputParser.
  • Integrated GPT-4 Omni LLM.
  • Enabled support to bind the backend server on any address.
  • Added the "q" query parameter to enable use as a search engine.

• Improvements

  • Updated the UI of the rewrite button.

Full Changelog: v1.4.0...v1.5.0

v1.4.0

v1.4.0

What's Changed

• Use value instead of selected props in <select>.
• Add support for local embedding models.
• Use Debian based Node image for TransformerJS.
• Remove expose directive in Docker compose file.

New Contributors

• @WanQuanXie made their first contribution in #79

Full Changelog: v1.3.4...v1.4.0

v1.3.4

v1.3.4

Changes

• Added better error handling for WebSocket handlers.

Fixes

• Model provider state having no values even if there are available models.

Full Changelog: v1.3.3...v1.3.4

v1.3.3

v1.3.3

Fixes

• Invalid provider, no models available and select not working / getting visible.

Full Changelog: v1.3.2...v1.3.3

v1.3.2

v1.3.2

Fixes

• Invalid provider when trying to load embedding models and chat models

Full Changelog: v1.3.1...v1.3.2