github valentinfrlch/ha-llmvision v1.3.7
AWS Bedrock, reconfigure providers and bug fixes
on GitHub

latest releases: v1.6.0-alpha.1, v1.5.2, v1.5.2-rc.1...
9 months ago

1.3.7 Release Notes

⚠️ Breaking Changes

Warning

If you were using any custom OpenAI-compatible providers, you'll need to set them up again.

Tip

The blueprint needs to be updated separately. You can update it by reimporting.
Go to Settings > Automations & scenes > Blueprints > (...) > Re-import blueprint

Features

Integration

Blueprint

  • Add filter for Frigate zones and option to ignore stationary objects by @amcfague

Bug Fixes

Integration

  • Fixed an issue preventing proper setup of some custom OpenAI-compatible providers (#173) by @bill9109 and @Smiley73
  • Resolved a setup failure that occurred when "Event Calendar" was selected as the first provider (#176) by @valentinfrlch
  • Fixed duplicate API calls in the Ollama provider by @Smiley73
  • Fixed a bug where stream_analyzerand video_analyzer could fail if stream resolution changed during recording (#158, #142) by @valentinfrlch
  • Increased max_tokens for event titles to prevent truncation (#182)
  • Translation improvements

Blueprint

  • Fixed a bug that caused the blueprint to fail when MQTT was not configured (#155) by @valentinfrlch
  • Fixed a bug that where the blueprint would fail when motion sensors contained "camera" in their name (#101) by @valentinfrlch
Check out latest releases or
releases around valentinfrlch/ha-llmvision v1.3.7

Don't miss a new ha-llmvision release

NewReleases is sending notifications on new releases.

Get notifications