How to use this with vLLM?

Hi @fh2019ustc and @xfey and Team,

How should I deploy this model with vLLM? Is it just `vllm serve "ByteDance/Dolphin"` ?

And then how to call the server for page-level and element-level parsing?

Can you provide demo code?

Thanks.

No file or test is named. Review the repository's model-serving and page- and element-level parsing entry points, then verify the vLLM command and example calls; done means the repository contains demo guidance for deploying the model and calling both parsing modes.

Written by the indexing model from the issue text.