[ENH] Add support for Ollama assistants

[smokestacklightnin]

This PR adds support for Ollama assistants.

[pmeier]

@nenb This violates "same response schema" part of #425 (comment) 😖

[pmeier]

@nenb As follow-up to #376 (comment), I've refactored the logic we merged in #425 a little to make it even more flexible.

• We now have an HttpApiCaller base class that handles the different streaming protocols for us. It is instantiated by the base class of the builtin assistants and parametrized by the streaming protocol, which is set as class constant.
• This new HttpApiCaller also handles JSON streaming for Google assistants now and thus abstracts the complexity away.
• I've renamed the OpenaiCompliantHttpApiAssistant to OpenaiLikeHttpApiAssistant, because the term "compliant" is even less meaningless when adding Ollama into the mix
• Each subclass of OpenaiLikeHttpApiAssistant now has to implement answer itself given that their is no standard way of handling it. The only common functionality is the call to the API, which so far seems to be rather constant across all models.

[nenb]

This looks fine to me.

I've pulled the branch locally and confirmed that it works for OpenAI models and for a Llamafile that I had locally.