[Feature Request] Qwen3-VL architecture performance optimization #5928
Description
What is your request?
I am working on serving the qwen3-vl-8b-it model on max to compare with vLLM serving performance.
Does max has optimised for this?
{{ message }}
I am working on serving the qwen3-vl-8b-it model on max to compare with vLLM serving performance.
Does max has optimised for this?