How to use Danrisi/Canon_UltraReal with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Danrisi/Canon_UltraReal", torch_dtype=torch.bfloat16, device_map="cuda") prompt = "c2n0n, dslr photo, amateur photo, candid photo, available light, underexposed, high resolution. A d.va from overwatch walks along a grassy roadside while looking down at a smartphone in her right hand. She wears her in-game clothes, headset, whiskers markings, dark hair with bangs. Her face is softly focused and partly shadowed, with a neutral expression and lowered eyes. Dense dark green trees and shrubs fill most of the horizontal frame, while a weathered metal utility pole stands near the right edge. A passing bmw car is heavily blurred across the lower-left foreground, unstaged street-photography feel. The high-resolution image retains substantial foliage detail, but the dim available light, deep shadows, soft subject focus, and large area of empty canopy weaken subject clarity and composition. medium quality" image = pipe(prompt).images[0]
