Expand the options in LocalAI configurations

[ha2]

Is your feature request related to a problem? Please describe.

Please expand the options in your LocalAI calls to allow us to configure more elements of the URL.

Currently, the Scheme, Port and API_endpoint are hardcoded in the /modules/Q03_localai_connector.sh file:

• line 292: http://"${LOCAL_AI_IP}":8080/v1/chat/completions
• line 419: http://"${LOCAL_AI_IP}":8080/v1/models

Describe the solution you'd like

1. It would be nice if the AI_OPTION=3 parameter could be set as a command line option and/or fixed in the /config/ai_config.env file.
2. In the /config/ai_config.env file contains the

• Schema (HTTP/HTTPS) -> LOCAL_AI_SCHEME="HTTPS"
• Port (0-65535) -> LOCAL_AI_PORT="8080"
• Path to API endpoint: -> LOCAL_AI_API_ENDPOINT="/v1"
  • https://HOSTNAME/v1 - works with the OpenAI libraries and any OpenAI-compatible tool
  • https://HOSTNAME/llm - works with Claude Code and other Anthropic tools

Describe alternatives you've considered
Instead, we could use an HTTP proxy.

Priority issue
Are you already a Sponsor? - [Y/N]
N

Additional context

[m-1-k-3]

Good thing! Could you give it a try and make a PR?