brianpetro / obsidian-smart-connections

Chat with your notes & see links to related content with AI embeddings. Use local models or 100+ via APIs like Claude, Gemini, ChatGPT & Llama 3

Home Page:https://smartconnections.app

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

latest version is easily stuck

linmaobang opened this issue · comments

I upgraded to the latest version today, but I found that obsidian is easily stuck, mainly because when embedding model (update notes), the CPU ratio will suddenly be much higher! I chose the jina8k version. I know you said before that this model is not very stable, but it has the best support for Chinese. I hope to enhance the stability of this model. Thank you for your time and creating such an amazing plugin!
屏幕截图 2024-05-16 212030

If you have not already tried using the Smart Connect software to improve local embeddings, it's worth trying to see if it helps with this https://github.com/brianpetro/smart-connect?tab=readme-ov-file#smart-connections-obsidian-plugin

You can also try reducing the "Local embeddings max tokens" setting 🌼