tensorrtllm_backend
tensorrtllm_backend copied to clipboard
Qwen2-VL support
the newest version support Qwen2-VL? especially the mrope param need to be sent to LLM