yisol
yisol
Hello, thanks for reporting this issues. Can you provide the original human image?
Hello, our model is based on [SDXL 1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0) and [SDXL inpainting 0.1](https://huggingface.co/diffusers/stable-diffusion-xl-1.0-inpainting-0.1).
Hello, you can use pull request for any suggestion for modification.
Hello, demo model is a try-on model for upper garments. It will automatically mask the upper body of image. You can uncheck the checkbox [Use auto-generated mask] and manually mask...
@ankitsiliconithub We will update the demo page to support lower body garments and dresses in the future. But you can try it now with the inference script for dresscode. (--category="lower_body")
We will update the demo page to support lower body garments and dresses in the future. But you can try it now with the inference script for dresscode. (--category="lower_body", --category="dresses")
Hello, we tested with torchvision=0.15.2=py310_cu118 as specified on environment.yaml file. You can check your torchvision library version.
Hello, you can run inference.py with your own dataset for batch generation. You have to prepare mask image and densepose image. You can refer [this link](https://blog.csdn.net/weixin_42815846/article/details/128510223) for preparing densepose image.
Hello, fixed the typo in the code. Thanks for reporting.
Hello, we used mixed precision(fp16) for inference. It should be worked with VRAM < 18GB on fp16. You can also try more optimizations like offloading @vcadillog mentioned for further reducing...