Open Issues Need Help
View All on GitHubAI Summary: The task is to add support for new embedding models to the Tezcat Obsidian plugin. This involves updating the plugin's code to include the specified Qwen3 (0.6B, 4B, 8B) and Gemini embedding models as selectable options for users. The implementation should consider both local (Ollama) and non-local (e.g., OpenAI API) model deployment scenarios. Thorough testing and documentation updates are also required.
Tezcat is an Obsidian plugin that uses AI embeddings to index your vaults and integrate search over your thoughts into your workflow.
AI Summary: The task requires scaling the scores produced by the hybrid search functionality in the Tezcat Obsidian plugin. Currently, hybrid search scores are significantly lower than vector similarity scores, making them less intuitive for users filtering search results. The solution involves adjusting the scoring mechanism, either during calculation or post-processing, to achieve a more balanced and user-friendly score range.
Tezcat is an Obsidian plugin that uses AI embeddings to index your vaults and integrate search over your thoughts into your workflow.