search

nerdyrodent / VQGAN-CLIP

Just playing with getting VQGAN+CLIP running locally, rather than having to use colab.
Other
2.59k stars 427 forks source link

Image prompt doesn't seem to work #74

Closed somecoulombs closed 2 years ago

somecoulombs commented 2 years ago

when I use the -ip flag to specify a starting image, I can see in the text output that it's using a text prompt as well as an image prompt, but the output preview always starts out as a brown blur, and not as the input image. The end result doesn't look anything like the image prompt, either

Is this a bug? Or am I doing something wrong

radugrecu97 commented 2 years ago

You're looking for -ii

vq_parser.add_argument("-ii", "--init_image", type=str, help="Initial image", default=None, dest='init_image')

I'm not sure how image_prompt is supposed to work though

somecoulombs commented 2 years ago

Ahhh, yep you're right, I was using -ip instead of -ii, thank you!