fancyfeast/llama-joycaption-beta-one-hf-llava base llava model?

[Passenger12138]

I am using VLLM offline batch inference to infer your model, and I would like to know which version of LLAVA is being used in this version of fancyfeast/llama-joycaption-beta-one-hf-llava. I can align the template.

[fpgaminer]

JoyCaption Beta One shares the same architecture as the original LLaVA model, but otherwise does not share any configuration/etc with it. If by template you mean the chat template: https://huggingface.co/fancyfeast/llama-joycaption-beta-one-hf-llava/blob/main/chat_template.json