selfhosted ollama server access

[Construkted-Reality]

Problem or use case
I don't run ollama on my local weak laptop, but I do have a beefy GPU server which is running ollama server, llama.cpp server and vllm server.

It would be great if selfhosted OpenAPI compliant API could be used for the post-processing.

Proposed solution
Expand the options in the configuration page (both cli and desktop app) to allow for either ollama or OpenAi compliant API access to a server IP and port of my choosing.

Alternatives considered
no

Additional context
no

[frankdierolf]

Same as in #16

After the next release, I think, I will make a fork for you with this patch. So I we can test it directly. I assume you are using Linux, right? Can you give me what you are running, then I can throw you the compiled version directly to you

Cheers, Frank ✌️

[Construkted-Reality]

running Mint Linux 22.2
Any package built for ubuntu will work