Instructions to use Boogu/Boogu-Image-0.1-Edit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Boogu/Boogu-Image-0.1-Edit with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Boogu/Boogu-Image-0.1-Edit", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
Edit?
No restoration of old photos.
Could you please provide some examples to show these failure cases (e.g., your input & prompt and the model's output)?
Hello dev,
Thanks for releasing Edit Turbo.
I tried the comfy version comparing to https://demo-edit-turbo-1k.boogu.org/.
and the result is quite different, but if I describe more edit instruction on what is missing in mine vs the web,
then the result is starting to get more similar to your web, so, maybe your online page has extra prompt enhancement?
what prompt enhancement should we use to match the result of web? thanks
Hi, @kfcpri . Many thanks for your support.
There's a tick box called "Enable Thinking" in our demo page, which will do prompt enhancement for the input instruction. But this option is turned off by default.
We are still optimizing our prompt enhancement, but in general, if you make your prompt more concrete, Boogu tends to perform better.
Could you give some cases to show your observations?