How does it score against a similar finetuned 3.5? The finetuning costs for 3.5 are negligible the finetuning of a llama is not. Is your plan to just provide th

How does it score against a similar finetuned 3.5? The finetuning costs for 3.5 are negligible the finetuning of a llama is not. Is your plan to just provide the finetuned model for their special use cases? Is there free prompting available?