Instructions to use baruga/hideous-blobfish with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use baruga/hideous-blobfish with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("baruga/hideous-blobfish", dtype=torch.bfloat16, device_map="cuda") prompt = "a photo of blofi fish in Times Square." image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
Description
This is a Stable Diffusion model fine-tuned on the infamous blobfish (often remarked as the ugliest animal in the world) for the DreamBooth Hackathon 🔥 animal theme. To participate or learn more, visit this page.
To generate blobfish images, use a photo of blofi fish in [your choice] or experiment with other variations. For some reason, GFC scale 5 seems to give the best results, at 7 images start get "overcooked". Despite multiple training runs with various settings, I couldn't fully solve this problem. Additional modifiers and negative prompts may also improve results.
Examples
a photo of blofi fish wearing a beautiful flower crown.
a photo of blofi fish in nerdy glasses.
a photo of blofi fish at the Arctic in a fluffy hat.
top rated surrealist painting of blofi fish by Salvador Dalí, intricate details.
top rated colorful origami photo of blofi fish.

Usage
from diffusers import StableDiffusionPipeline
pipeline = StableDiffusionPipeline.from_pretrained('baruga/hideous-blobfish')
image = pipeline().images[0]
image
- Downloads last month
- 6