[Feature] Support passing timeout parameter to LiteLLM

[markus-meingast]

I am attempting to run OpenKB using a local Ollama model as LLM backend:

language: en
model: ollama/<model>
pageindex_threshold: 20

For larger documents I am running into default (10min) timeout issues: litellm.Timeout: Connection timed out. Timeout passed=600.0. It would be great if timeouts could be set via the config.yml, e.g.:

language: en
model: ollama/<model>
pageindex_threshold: 20
timeout: 1200 # <--

[KylinMountain]

Thanks @markus-meingast — that's a reasonable ask. We'll add support for configuring the timeout. Will update here once it's in. 🙏

[KylinMountain]

Hi @markus-meingast — added in #134.

To try it, install from source and set timeout: (seconds) in your .openkb/config.yaml:

pip install git+https://github.com/VectifyAI/OpenKB.git

timeout: 1200

Thanks for the report! 🙏