import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("SamuelTallet/Krea-2-Turbo-SDNQ-3bit-dynamic-hadamard256", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]This is Krea 2 Turbo optimized using SDNQ with UINT3 dynamic quantization and Hadamard Rotation (Group size: 256).
Sample
Prompt:
A heavily burdened, blue-skinned warrior wanders through a starry dreamscape in a detailed 2D illustration merging sci-fi and feudal fantasy aesthetics. Viewed in profile facing left, the figure features mottled skin, a dark half-mask with gold accents, and an elongated blue headpiece. They wear a complex, patched tunic with a thick brown fur mantle, carrying two ornately hilted katanas at the hip. A towering pack strapped to their back is lashed with thick ropes, securing gourds, wrapped bundles, and swirling cyan vapor. The foreground holds pale blue plants against a gradient sky transitioning from deep indigo to bright cyan. White stars and a massive dark blue crescent shape frame the character, rendered with vivid cel-shaded coloring and intricate line art.
Seed: 42
Usage
Install Torch, Diffusers (Git), SDNQ 0.2.0 and Triton.
Use CFG 0 and 8 steps.
- Downloads last month
- 422