by nerijs
Open source · 6k downloads · 603 likes
Pixel Art XL is an AI model specialized in generating high-quality pixel art images. It excels at producing retro visuals with sharp details and vibrant colors, perfect for video games or artistic creations inspired by the 80s and 90s. Its main strength lies in its ability to generate perfectly proportioned pixelated images, even after significant resizing, thanks to optimized downscaling techniques. The model stands out for its ease of use, requiring neither trigger keywords nor specific styles to function effectively. It adapts seamlessly to both isometric perspectives and classic views, offering great flexibility for a variety of creative projects.
Downscale 8 times to get pixel perfect images (use Nearest Neighbors) Use a fixed VAE to avoid artifacts (0.9 or fp16 fix)
Use it with a LCM Lora!
Use 8 steps and guidance scale of 1.5 1.2 Lora strength for the Pixel Art XL works better
from diffusers import DiffusionPipeline, LCMScheduler
import torch
model_id = "stabilityai/stable-diffusion-xl-base-1.0"
lcm_lora_id = "latent-consistency/lcm-lora-sdxl"
pipe = DiffusionPipeline.from_pretrained(model_id, variant="fp16")
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
pipe.load_lora_weights(lcm_lora_id, adapter_name="lora")
pipe.load_lora_weights("./pixel-art-xl.safetensors", adapter_name="pixel")
pipe.set_adapters(["lora", "pixel"], adapter_weights=[1.0, 1.2])
pipe.to(device="cuda", dtype=torch.float16)
prompt = "pixel, a cute corgi"
negative_prompt = "3d render, realistic"
num_images = 9
for i in range(num_images):
img = pipe(
prompt=prompt,
negative_prompt=negative_prompt,
num_inference_steps=8,
guidance_scale=1.5,
).images[0]
img.save(f"lcm_lora_{i}.png")
Don't use refiner
Works great with only 1 text encoder
No style prompt required
No trigger keyword require
Works great with isometric and non-isometric
Works with 0.9 and 1.0
v1: Initial release