Zheng-Chong / CatVTON

CatVTON is a simple and efficient virtual try-on diffusion model with 1) Lightweight Network (899.06M parameters totally), 2) Parameter-Efficient Training (49.57M parameters trainable) and 3) Simplified Inference (< 8G VRAM for 1024X768 resolution).
Other
887 stars 104 forks source link

Error occurred when I tried to change input width and height! #34

Closed miniwa00 closed 2 months ago

miniwa00 commented 2 months ago

Hi! Thanks for your great work!

When I don't give any specific width and height value, It works perfectly But when I tried to change these width and height, the code runs successfully without any warnings or errors, but the result that I've got is just black empty image. (below image) I think resolutions were changed into given options.

Thanks!

image

Zheng-Chong commented 2 months ago

CatVTON's training does not currently accommodate various aspect ratios and resolutions; the 1024 model only produces satisfactory results at a resolution of 1024x768. However, adapting to variable resolutions is on our list of improvements for future development.

miniwa00 commented 2 months ago

I understand! Thanks for your quick answers ;)

miniwa00 commented 2 months ago

I wil close it! Thanks again!