SGLang Deployment approach

[Mattz-CE]

Great work on inference server setup! I'm wondering if SGLang deployment approach has been experimented or not, as it has became an staple for vision models.

[boyugou]

Hi! I'm not pretty sure about that. I just search a little bit about "Qwen2-VL SGLang". Maybe you can check something like this:

sgl-project/sglang#1546

[boyugou]

Also see:

https://docs.sglang.ai/references/supported_models.html

https://modal.com/docs/examples/sgl_vlm

[Mattz-CE]

Thanks! I'm assuming the setup process would be similar to the instructions for Qwen 2?