Downloads · 30 days
26
41% of all-time downloads
MikeChandler/mike-krea2
mike-krea2 is a text-to-image model from MikeChandler. Use it when you need an image from a text prompt. It is set up for diffusers. The card lists the license as apache-2.0.
Downloads · 30 days
26
41% of all-time downloads
All-time downloads
64
Public
Repo size
778 MB
Likes
0
Public
Click a slice to open those files.
.safetensors576 MB · 59%
From the Hugging Face model README
A DreamBooth-LoRA for Krea 2, trained on Krea 2 RAW and shown on Krea 2 Turbo. The samples below were generated with this LoRA on Turbo (8 steps).
Use the token mike_krea to invoke the concept.

"A futuristic cyborg samurai standing atop a neon-lit skyscraper in a rainy cyberpunk Tokyo, cinematic lighting, mike_krea."

"A whimsical tiny mouse wearing a royal velvet cape and crown, sitting on a giant mushroom in a glowing enchanted forest, mike_krea."

"An astronaut floating in a surreal cosmic ocean where planets are iridescent bubbles and jellyfish drift through the stars, mike_krea."
import torch
from diffusers import Krea2Pipeline
pipe = Krea2Pipeline.from_pretrained("krea/Krea-2-Turbo", torch_dtype=torch.bfloat16).to("cuda")
pipe.load_lora_weights("MikeChandler/mike-krea2")
image = pipe("A futuristic cyborg samurai standing atop a neon-lit skyscraper in a rainy cyberpunk Tokyo, cinematic lighting, mike_krea.", num_inference_steps=8, guidance_scale=0.0).images[0]
image.save("output.png")