How to use numz/SeedVR2_comfyUI with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("numz/SeedVR2_comfyUI", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0]
In the case of special images, the model produces misaligned enlarged images, likely due to insufficient consideration during the image block composition process.
No matter how the enlargement size is adjusted, the red-tipped area in the attached image does not improve.
You may want to discuss your problems here.
· Sign up or log in to comment