import torch
from diffusers import DiffusionPipeline
from diffusers.utils import load_image
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("GreeneryScenery/SheepsControlV3", dtype=torch.bfloat16, device_map="cuda")
prompt = "Turn this cat into a dog"
input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png")
image = pipe(image=input_image, prompt=prompt).images[0]V3
1 epoch. 🤗 Honestly, V2 might be better. Try it here.
Flaws:
- The conditioning image does not really affect the output image
Things to improve:
- More epochs
- Better prompts
- More data
- Preprocess the data
Examples (Cherry-picked)
- Conditioning image:
Images:
A pig
Cow, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
Donkey, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k.png
Goat, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
Horse
Sheep, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
- Conditioning image:
Images:
Fire hydrant, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
Fire hydrant
Red fire hydrant with white background
Red fire hydrant
- Conditioning image:
Images:
A bicycle
Bicycle, blue
Bicycle, realistic
Bicycle, red
Bicycle, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
Bicycle
- Conditioning image:
Images:
Airplane, sideview, top
Airplane, trending on artstation, artstationHD, artstationHQ, patreon, 4k, 8k
- Downloads last month
- 6