More than 32 exposed entities?

[starman2k01]

I installed and had 530 entities already exposed, causing integration to error out. Is there a model that will take much more entities?

Also I get an error "I'm sorry, I didn't produce a correctly formatted tool call! Please see the logs for more info." but there are no log entries under Settings->System-> Logs. Where are the logs for the home-llm?

[acon96]

To get the logs to be printed, you need to enable debug logging for the integration: https://www.home-assistant.io/docs/configuration/troubleshooting/#enabling-debug-logging

To handle that many entities, you probably need to have at least 128k context size. Phi 3 mini, small, & medium should support that context size.

[starman2k01]

Thank you; I've been probing reddit to find a small model that might work. Everyone says I should use RAG and not feed my entities into context.
Also I deleted the integration in order to load phi 3 mini, but it get this error:
"Pip returned an error while installing the wheel! Please check the Home Assistant logs for more details."
It seems once you install local LLM, you cannot re-install it?

[starman2k01]

Detail error logs:
Unable to install package https://github.com/acon96/home-llm/releases/download/v0.3.6/llama_cpp_python-0.2.88-cp313-cp313-linux_x86_64.whl: × Failed to download llama-cpp-python @ │ https://github.com/acon96/home-llm/releases/download/v0.3.6/llama_cpp_python-0.2.88-cp313-cp313-linux_x86_64.whl ├─▶ Failed to fetch: │ https://github.com/acon96/home-llm/releases/download/v0.3.6/llama_cpp_python-0.2.88-cp313-cp313-linux_x86_64.whl ╰─▶ HTTP status client error (404 Not Found) for url (https://github.com/acon96/home-llm/releases/download/v0.3.6/llama_cpp_python-0.2.88-cp313-cp313-linux_x86_64.whl)

Followed by:
Error installing llama-cpp-python. Could not install the binary wheels from GitHub for platform: x86_64, python version: 3.13. Please manually build or download the wheels and place them in the /config/custom_components/llama_conversation directory.Make sure that you download the correct .whl file for your platform and python version from the GitHub releases page.

[starman2k01]

BTW your integration is a must-have for future HA use. It's fast and it works!

[acon96]

It looks like Home Assistant moved to Python 3.13. I'll need to push a new version with wheels compatible with that runtime.

[starman2k01]

I deleted the llama_conversation folder and now re-installing just spins forever trying to install llama.cpp.