wan-image-video-gen-edit
SKILL.md
Wan Models
Wan Models, created by Alibaba Group, are popular image and video generation and editting models and widely adopted around the world. This skill integrates with Wan Modles APIs on ModelStudio(Bailian-Alibaba Model Service Platform).
text2image generation
Gen images from text prompt
python3 {baseDir}/scripts/wan-magic.py text2image --prompt "一个女生站在楼顶的阳台上，夕阳照在她的脸上"
python3 {baseDir}/scripts/wan-magic.py text2image --prompt "一位长发女孩坐在书桌前，背对着镜头，戴着耳机。阳光透过窗户洒进房间，照亮了她和周围散落的书籍与杂物" --size 1280*1280
python3 {baseDir}/scripts/wan-magic.py text2image --prompt "女生优雅地倚在车门旁，身穿红色褶皱长裙，在复古色调的室内场景中缓慢转身看向镜头，霓虹光斑在玻璃窗上流动，轻微晃动，背景家具逐渐虚化凸显人物独白，画面带有电影胶片颗粒质感，港风朦胧光影映照出淡淡的忧伤情绪" --quantity 1
Options
--quantity: Number of images (default: 1, max: 4)--prompt: User Prompt for Image Generation--size: Image resolution(default:12801280,support resolutions with a width and height from 512 to 1440 pixels, provided the total pixel count does not exceed 14401440. Common resolutions:12801280,11041472,14721104,9601696,1696*960)