VLLM 部署多模态模型Qwen3.8-27B-FP8处理图片和视频报错，接入deepseek-harness-desktop，上传图片和视频报错： model "Qwen3.8-27B-FP8" does not declare image input; switch to an image-capable model to read images
VLLM 部署多模态模型Qwen3.8-27B-FP8处理图片和视频报错，接入deepseek-harness-desktop，上传图片和视频报错：
model "Qwen3.8-27B-FP8" does not declare image input; switch to an image-capable model to read images