Instructions to use FireRedTeam/FireRed-Image-Edit-1.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use FireRedTeam/FireRed-Image-Edit-1.1 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("FireRedTeam/FireRed-Image-Edit-1.1", 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] - Notebooks
- Google Colab
- Kaggle
Interesting Behavior when Already Edited Image used
#10
by gjnave - opened
Hey guys, just poihnting out of curiosity the following:
if you throw in a previously not edited image, the Firered model will edit it perfectly (great job!)
the intersting behavior comes when you throw in a previously edited image then the background is all messed up. Almost as if fire red does something to the image during the original edit that interferes with the second run.
I tried this on several and had the same experience. Just curious
thanks!
watermark gone wrong? That is interesting. Because it shouldn't be a thing.
try to add a fine layer of grain/noise and redo the test to see if the embedded watermark is gone ;-)