[Feature]: support Mistral-Large-Instruct-2407 function calling

[ybdesire]

🚀 The feature, motivation and pitch

1. Mistral-Large-Instruct-2407 performance is great, even Tool Use & Function Calling : https://mistral.ai/news/mistral-large-2407/
2. the feature requirement is(maybe) similar as : [Feature]: VLLM suport for function calling in Mistral-7B-Instruct-v0.3 #5156

Alternatives

No response

Additional context

No response

[BUJIDAOVS]

Same request! I really need vLLM to support function call API interfaces. In fact, many of the currently released LLMs already support function call capabilities. This is very important for downstream application development. Plz

[K-Mistele]

if it uses the same function call format as Mistral 7B v0.3 it should be supported in #5649

[Sapessii]

@K-Mistele first of all thank you and good job on what you're doing to implement FC on vllm.

I'm using a quantized version of mistral large 2407, don't know if the template is the same but you can check it on line 6176 here https://huggingface.co/ModelCloud/Mistral-Large-Instruct-2407-gptq-4bit/blob/main/tokenizer_config.json

[github-actions]

This issue has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this issue should remain open. Thank you!

[flefevre]

MistralAI highligth the compatibility between vllm and mistral-small https://huggingface.co/mistralai/Mistral-Small-Instruct-2409 but baddly there are several tickets on the fact Vllm do not implement "function calling" aka tools for Mistral-Small.
Could you confirm the statut of supporting function calling for mistral models such as small, large ?
And could you share the command to start vllm with tools enable?