윈디하나의 누리사랑방. 이런 저런 얘기
Traceback (most recent call last):컴퓨터로 부동소수점 연산하다보면 NaN 이 생길 수 있기 때문에 이부분에 대해서는 대책을 세워야 한다. SD에서는 아래와 같이 해보라고 한다.
File "/home/windy/stable-diffusion-webui/modules/call_queue.py", line 56, in f
res = list(func(*args, **kwargs))
File "/home/windy/stable-diffusion-webui/modules/call_queue.py", line 37, in f
res = func(*args, **kwargs)
File "/home/windy/stable-diffusion-webui/modules/txt2img.py", line 56, in txt2img
processed = process_images(p)
File "/home/windy/stable-diffusion-webui/modules/processing.py", line 503, in process_images
res = process_images_inner(p)
File "/home/windy/stable-diffusion-webui/modules/processing.py", line 657, in process_images_inner
devices.test_for_nans(x, "vae")
File "/home/windy/stable-diffusion-webui/modules/devices.py", line 152, in test_for_nans
raise NansException(message)
modules.devices.NansException: A tensor with all NaNs was produced in VAE. This could be because there's not enough precision to represent the picture. Try adding --no-half-vae commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.