UmeAiRT/Ume-Modern_Pixel_Art
Viewer β’ Updated β’ 100 β’ 13
How to use UmeAiRT/FLUX.1-dev-LoRA-Modern_Pixel_art with Diffusers:
pip install -U diffusers transformers accelerate
import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("UmeAiRT/FLUX.1-dev-LoRA-Modern_Pixel_art")
prompt = "(ruins, broken, summer, dappled sunlight:1.2), sun, tree, forest, scenery, rock, reflection, water, ancient, overgrown, mountains, sunset, clouds, mountainous horizon, fantasy, medieval, 1other, looking away, cloak, (pixel art, pixelated:1.2), (masterpiece, exceptional, best aesthetic, best quality, masterpiece, extremely detailed:1.2), in the style of umempart"
image = pipe(prompt).images[0]


Here's a new LoRA in a much more modern style.
Based on games I've personally really enjoyed from the modern 2d pixel art category.
This is the first time I've worked with such a large dataset (100), so I'm not sure my settings are optimal.
Don't hesitate to send me feedback or suggestions!
You should use umempart to trigger the image generation.
Weights for this model are available in Safetensors format.
Download them in the Files & versions tab.
epochs : 20
steps : 2000
lora_rank : 32
optimizer : prodigy
resolution : 1024
Base model
black-forest-labs/FLUX.1-dev