darhsu/rubbrband

Rapidly experiment with open-source machine learning models

Rubbrband: Monitoring for Stable Diffusion

Rubbrband detects deformities in your images generated by Stable Diffusion at scale.

Installation

pip install rubbrband

Usage

import rubbrband
rubbrband.init("YOUR_API_KEY")
rubbrband.upload(image="https://example.com/image.png", prompt="Prompt used to generate image")

Uploading Images to Rubbrband

1. Using a URL

rubbrband.upload(image="https://example.com/image.png", prompt="Prompt used to generate image")

2. Using a PIL Image

from PIL import Image
rubbrband.upload(image=Image.open("/path/to/image.png"), prompt="Prompt used to generate image")

3. Using a path

rubbrband.upload(image="/path/to/image.png", prompt="Prompt used to generate image")

4. Using a context manager

with open("/path/to/image.png", "rb") as f:
    rubbrband.upload(image=f, prompt="Prompt used to generate image")

Getting Evaluation Results

filename = rubbrband.upload(image="/path/to/image.png", prompt="Prompt used to generate image")
evaluation = rubbrband.get_image_metadata(filename)