Skip to content

Dataset · Chan-Y

Synthetic pixel art dataset

Chan-Y/pixelart-308k

A dataset of 308K synthetic pixel art images paired with text descriptions.

What's inside

Contains 308,765 256x256 RGB pixel art style images created via a two-stage pipeline using Google Gemma models for prompts and FLUX.2-klein-4B for image generation. Each image is provided with a detailed text description.

What you can do with it

  • Training image-to-text and pixel art generation models
  • Fine-tuning LoRA for image generators in a specific style
  • Researching synthetic datasets and generation pipelines
Tasks
Text → image, Image → text
Size
100K–1M rows
Downloads this month
842
Likes
10

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Chan-Y/pixelart-308k")