chore: implement local ai config

[appflowy]

Model Selection Process

1. Collect Available Models:

   • Retrieve server models (marked as non-local).
   • Add the local AI model if it’s enabled.

2. Determine the Active Model:

   • Get the global active model from workspace settings.
   • Retrieve the user-selected model from stored preferences using a key derived from the source.
     • If no preference is found and a local model is available, choose the local model.
     • Otherwise, use the stored user preference.
   • If the chosen model isn’t in the available list, fall back to the global active model.