Hugging Face's logo

Datasets:
jsalazar
/
sora

Modalities:
Image
Dataset card Data Studio Files

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

Sora Image Dataset

BUILDING DATASET, CURRENTLY UNAVAILABLE

This dataset contains images generated using OpenAI's Sora text-to-image model. The images are organized in a structured way to facilitate access and analysis.

Dataset Structure

The dataset is organized as follows:

/
├── sora_global.csv           # Complete dataset information
├── images_part_001.zip      # First 10,000 images
├── images_part_002.zip      # Next 10,000 images
└── ...

Each zip file contains:

  • Up to 10,000 images.
  • A CSV file with the same name as the zip (e.g., images_part_001.csv) containing information for the images in that zip.

Image Naming Convention

Each image is named using its unique ID from the Sora generation process, followed by the .png extension. For example:

gen_01jqe558kpf43shtzrhzdnh4hy.png

CSV Structure

The CSV files contain the following columns:

Column Description Example
ID Unique identifier for the generated image gen_01jqe558kpf43shtzrhzdnh4hy
Task ID Identifier for the generation task task_01jqe3bkb3f5gt6w9m9jvm2w06
Like Count Number of likes received 0
Created At Timestamp of image creation 2025-03-28T10:35:58.133166Z
Title Title given to the generation Vintage Hacker Scene
Prompt Text prompt used to generate the image "gruny analog image of a hacker wearing Mr robot mask..."
Seed Random seed used for generation 12345
Width Image width in pixels 1024
Height Image height in pixels 1536
URL Temporary URL to the original image https://videos.openai.com/...

Usage

  1. The sora_global.csv file contains information for all images in the dataset.
  2. Each zip file contains up to 10,000 images and their corresponding information.
  3. The images in each ZIP file are ordered by date of acquisition.
  4. Each zip's CSV contains only the information for the images within that zip.

Notes

  • Image URLs in the CSV are temporary and may expire.
  • The dataset is organized in batches of 10,000 images to facilitate downloading and processing.
  • Each zip file is self-contained with both images and their information.

License

This dataset is made available for archival and transparency purposes only.
You are not permitted to download, use, modify, distribute, or reproduce any part of this dataset without the explicit permission of Lawwwing.

All rights reserved © Lawwwing, 2025.

Downloads last month
4