How can we cancel currently running generation for Wan 2.1?

Here your example code

https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/wanvideo/wan_1.3b_text_to_video.py

video = pipe(
    prompt="纪实摄影风格画面，一只活泼的小狗在绿茵茵的草地上迅速奔跑。小狗毛色棕黄，两只耳朵立起，神情专注而欢快。阳光洒在它身上，使得毛发看上去格外柔软而闪亮。背景是一片开阔的草地，偶尔点缀着几朵野花，远处隐约可见蓝天和几片白云。透视感鲜明，捕捉小狗奔跑时的动感和四周草地的生机。中景侧面移动视角。",
    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
    num_inference_steps=50,
    seed=0, tiled=True
)

How can this command be cancelled any way? I mean lets say 10 steps done and I want to cancel the generation - it doesnt have to return back a video but if returns even better

Start with examples/wanvideo/wan_1.3b_text_to_video.py and inspect the pipe(...) generation call shown in the issue. Reproduce a run, then trace how the 50 inference steps are executed to identify where cancellation could be handled. Done means a running generation can be cancelled after several steps without requiring a returned video.