README.md

---
license: gpl-3.0
task_categories:
- text-classification
language:
- en
size_categories:
- n<1K
tags:
- art
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype: string
  - name: sprite_id
    dtype: string
  splits:
  - name: train
    num_bytes: 2723127.984
    num_examples: 1232
  download_size: 1807666
  dataset_size: 2723127.984
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
# My Spraix Dataset

This dataset is a demo created for the Animator2D-v1.0.0 project, with the goal of generating pixel-art sprite animations from textual prompts. It is still in the preparation phase and contains sprite frames extracted and organized into subfolders, each with associated metadata.

## Dataset Details
- **Format**: PNG for frames, JSON for metadata.
- **Structure**: Numbered subfolders (e.g., `152_frames`) containing sequential frames (e.g., `frame_000.png`).
- **Usage**: Designed for training 2D animation generation models.

## Citation
This dataset was derived and inspired by the `pawkanarek/spraix_1024` dataset available on Hugging Face. Please cite the original dataset if you use this work:

- **Original Dataset**: `pawkanarek/spraix_1024`  
  Available at: [https://huggingface.co/datasets/pawkanarek/spraix_1024](https://huggingface.co/datasets/pawkanarek/spraix_1024)