ConnectionError: Failed to connect to API: timed out

[kanwarplaha]

No matter which model I choose and what operation I try from my notebook with two sources (one web, one local document), I get this error:

ConnectionError: Failed to connect to API: timed out

Local ollama is working finr with Open WebUI for queries. However, from OpenNotebook, I keep getting timed out error.

Is there any reliable way to eliminate these errors and get this working? Thank you.

[AnnaChen1228]

I also have this problem in upload notebook when using local models (Llama3 70B, BGE-Large via Ollama), but using GPT works fine. Other transformation functions work correctly, only the insight function has issues. I've tried increasing the timeout settings, but it didn't help.

Is there any way to solve this timeout issue with local models? Thanks.

[chariraghavann-svg]

I also have the same problem while using google gemini's text embedding 004

[PedroNeto05]

I have the same issue

[g0ldingkev]

i have the same issue. GPU consumption increases for a few seconds then craps out, only happens when loading a file though, using text input seems to be fine. Ollama Model is oss:20b, and Deepseek:14b. but it seems that the processing actually works because you can chat to the doc

[lfnovo]

Guys, this seems to be different issues. For Ollama, anybody having trouble with it please share your env file. Also, look at the following new doc page that details the 3 types of Ollama installations. - https://github.com/lfnovo/open-notebook/blob/main/docs/features/ollama.md

This should help @kanwarplaha , @AnnaChen1228 and @g0ldingkev

As for @chariraghavann-svg's issue, this is a different issue as Gemini <> Ollama. I am opening a new ticket for this. See #129