FIM from OpenWebui seems impossible to make work

[mtoniott]

Hey !

Thanks for the project. I really would love to use it, but I have problems making it work on my machine.

I have a server dedicated to AI workloads, where I use llama.cpp to server to open webui the models. I don't want to expose directly llama.cpp endpoint and prefer to go through open webui for security.

I want to make FIM work on twinny on my laptop. Chat model work like a charm with "/api/v1/" api path.

for FIM, I really have no clue what is wrong. VSCode threw me an error 405 when I used /api/v1/ , so I tried different endpoint and the only different result that I get is error 400 when using "/api/v1/chat/completions".

Here is a picture from my setup. What's wrong ?

This issue (#224) seemed to indicate that I should use /ollama/api/generate, which also gives me a nice error 400.

What should I do / where does the problem comes from ?