move RuntimeError if no provider is given

[amaloney]

Moving the RuntimeError closer to when the provider is defined from the args object. I found this error message made a lot more sense than the one found in the try-block when/if I tried to run lumen-ai on the command line with no provider set.

resolves: #1052

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.

Project coverage is 57.62%. Comparing base (c2404e1) to head (6d76945).

Files with missing lines	Patch %	Lines
lumen/command/ai.py	0.00%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1053   +/-   ##
=======================================
  Coverage   57.62%   57.62%           
=======================================
  Files         109      109           
  Lines       14228    14228           
=======================================
  Hits         8199     8199           
  Misses       6029     6029           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ahuang11]

Thanks for pointing that out.

The reason why it's below is because of detect_provider().

        if llm_model_url and provider and provider != "llama":
            raise ValueError(
                f"Cannot specify both --llm-model-url and --provider {provider!r}. "
                f"Use --llm-model-url to load a model from HuggingFace."
            )
        elif llm_model_url:
            provider = "llama"
        elif not provider:
            provider = LLMConfig.detect_provider()

I think you can just replace this with the more helpful message?

                f"Could not find LLM Provider {provider!r}, valid providers include: {list(LLM_PROVIDERS)}."
            )