3D-Animation-Style / README.md
SOLRICKS's picture
Update README.md
6ed7ff8 verified
|
Raw
History Blame Contribute Delete
2.45 kB
metadata
license: cc-by-4.0
task_categories:
  - text-to-image
  - image-to-image
language:
  - en
tags:
  - 3d-animation
  - cinematic
  - stylized
  - synthetic
  - image-captioning
  - lora
  - solricks
pretty_name: 3D Animation Style
size_categories:
  - n<1K
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-00000-of-00001.parquet

3D Animation Style

A curated collection of 54 high-resolution synthetic images by SOLRICKS, designed around a warm, cinematic 3D animation aesthetic. The dataset combines richly lit environments, fantasy interiors, stylized animal characters and expressive human characters.

Dataset details

  • Images: 54 PNG files
  • Resolution: 50 images at 1254×1254 and 4 images at 1536×1024
  • Captions: English, manually curated
  • Trigger token: 3dsrx
  • Columns: image, text
  • Content: environments, interiors, animal characters and human characters
  • Data origin: synthetically generated and curated by SOLRICKS

Repository structure

3D-Animation-Style/
├── README.md
├── data/
│   └── train-00000-of-00001.parquet
├── metadata.csv
└── train.zip
    ├── 0001.png
    └── ... 0054.png

The Image Parquet file powers the Dataset Viewer with compact thumbnails and visible text captions. The train.zip and metadata.csv files provide the same source dataset in ImageFolder format for direct download and LoRA training workflows.

Loading the dataset

from datasets import load_dataset

dataset = load_dataset("SOLRICKS/3D-Animation-Style", split="train")
print(dataset[0]["image"])
print(dataset[0]["text"])

Intended uses

  • Training or fine-tuning 3D animation style LoRAs
  • Text-to-image and image-to-image experimentation
  • Image-captioning and visual style research
  • Creative concept development and reference-driven workflows

Limitations

This is a small, curated synthetic dataset and does not represent the full diversity of real-world subjects or visual styles. Generated scenes may contain minor stylization artifacts, simplified anatomy or imperfect decorative text. Human characters are fictional and are not intended to represent real people.

License

Released under the Creative Commons Attribution 4.0 International License. Attribution to SOLRICKS is required when redistributing or adapting the dataset.