______  ______   __       __  __   ______   ______   __  __   ______       __   ______    
/\  == \/\  __ \ /\ \     /\ \_\ \ /\  ___\ /\  == \ /\ \/\ \ /\___  \     /\ \ /\  __ \   
\ \  _-/\ \ \/\ \\ \ \____\ \____ \\ \ \____\ \  __< \ \ \_\ \\/_/  /__    \ \ \\ \ \/\ \  
 \ \_\   \ \_____\\ \_____\\/\_____\\ \_____\\ \_\ \_\\ \_____\ /\_____\    \ \_\\ \_____\ 
  \/_/    \/_____/ \/_____/ \/_____/ \/_____/ \/_/ /_/ \/_____/ \/_____/     \/_/ \/_____/ 
                                                                                                                                                                                                  

pixelo-flux

Prompt

better pixel, A pixelated image of a blue man and a white dog on a pink background made up of small square tiles. The blue mans head is on the left side of the image, while the white dog is to the right of the blue man. Both of the dogs are facing each other. The dogs mouth is slightly open, and the dogs tongue is sticking out of its mouth. The pink tiles are arranged in a grid pattern.

Prompt

better pixel, Captured at eye-level, a mosaic of a mans face is visible. The mosaic is made up of small square tiles, each with a different shades of gray and white. The mans eyes are squinted, and he is facing the right side of the frame. His body is positioned to the left of the image, and his hands are positioned in front of his body. The background is a combination of yellow and white squares, creating a grid effect. The image is taken from a low-angle perspective.

Prompt

better pixel, A pixelated image of a pink teddy bears face is seen on a white tiled surface. The face is made up of small pink squares, and black squares are arranged in a grid pattern. The eyes, ears, and nose of the bears head are also visible, and there is a brown stripe at the bottom of the face. The background is white, and the image is taken from a low-angle perspective.

---

Model description for pixelo-flux

Image Processing Parameters

Parameter	Value	Parameter	Value
LR Scheduler	constant	Noise Offset	0.03
Optimizer	AdamW	Multires Noise Discount	0.1
Network Dim	64	Multires Noise Iterations	10
Network Alpha	32	Repeat & Steps	20 & 2600
Epoch	14	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 24

Best Dimensions & Inference

Dimensions	Aspect Ratio	Recommendation
1280 x 832	3:2	Best
1024 x 1024	1:1	Default

Inference Range

• Recommended Inference Steps: 30–35

Setting Up

import torch
from pipelines import DiffusionPipeline

base_model = "black-forest-labs/FLUX.1-dev"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "Polycruz9/pixelo-flux"
trigger_word = "better pixel"  
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)

Trigger words

You should use better pixel to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.