The generated image contains noticeable input image artifacts.
#15
by
liqi68
- opened
The generated image contains noticeable input image artifacts.
How can we better generate novel views, and why does the model produce artifacts from the reference image?
same question 🤔
This is probably due to the version of diffusers. I install the newest version of diffusers, the problem is gone.
pip install git+https://github.com/huggingface/diffusers
I met the same question. Did you run the code with "QwenImageEditPlusPipeline.load_lora_weights"?