Local Custom Models

[sfxworks]

Models is currently enum without the ability to set a custom model name

I am running a local-ai deployment that I'd like to be able to specify custom models for. I have deepseek-coder running via a compatible openai endpoint.

Currently I have to rename the coder model to gpt-4 but id like it if I could just add custom models here like that and wizard-coder.

[trymeouteh]

Privy, which is a fork of Rubberduck supports Ollama models which would be cool to see in rubberduck