Closed eimantasurbonas closed 1 week ago
Currently there is an arbitrary limit of 10, this is something that will be improved 🌴
Currently there is an arbitrary limit of 10, this is something that will be improved 🌴
That's very dissapointing. Is there some sort of workaround atleast for now or some other obsidian plugin?
Supporters will get early access to the next version of Smart Chat in the near future. Unfortunately the existing version is too cumbersome to build/improve, so I've been working on a rewrite that will allow many improvements.
While I can't recommend alternatives to Smart Connections, primarily because I'm too focused on building what I see as the future of software rather than evaluating alternatives, I'm sure that is something ChatGPT could help you with 😉 🌴
Supporters will get early access to the next version of Smart Chat in the near future. Unfortunately the existing version is too cumbersome to build/improve, so I've been working on a rewrite that will allow many improvements.
While I can't recommend alternatives to Smart Connections, primarily because I'm too focused on building what I see as the future of software rather than evaluating alternatives, I'm sure that is something ChatGPT could help you with 😉 🌴
You mean its possible to build a custom bot in chatgpt and feed a few k of notes into it?
This GitHub issues section discusses software issues. While I wish I could advise everyone on how to empower themselves with AI, I am doing the best I can by developing open-source tools design for individuals and open-source components that other developers can use to build, too.
Also, there is no need to full quote my previous message in a one-on-one thread like this. What you're responding to is pretty clear without the quote.
Lastly, there is a link to join our Community Lean Coffee meetings in the Smart Connections settings. Bring your questions, and everyone has a chance to vote on what topics we discuss (next one is tomorrow ☕️)
🌴
Thanks for information. However I believe that if someone will search for related issue it would be great to provide atleast a temporary solution :)
So, if you are reading this and you're like me and need a workaround to use AI for all your notes, try Google NotebookLM (supports 50 data sources per notebook, with up to 500k words each and fast AI) or fabric.so.
Hey guys,
When using chat (with gemini-flash model and bge-micro-v2 embedding model) I noticed that gemini was not aware of some parts of context, so I started testing and it seems that Gemini is able to see only 10 notes at once.
So basically I have here 19 notes as shown in the explorer and in "Smart connections files" I can see them but gemini can see only 10 at once, even when using "Based on my notes" prompt.
How to fix this? Do I need a paid API or to become a supporter or something else?
This is very frustrating. I tried various models/prompts...