| --- |
| license: cc-by-4.0 |
| task_categories: |
| - image-to-text |
| - text-to-image |
| language: |
| - en |
| tags: |
| - laion |
| - recaptioned |
| - vision-language-alignment |
| - clip-training |
| - webdataset |
| pretty_name: Recaptioned LAION (Gemma-3-27B) |
| size_categories: |
| - 100K<n<1M |
| --- |
| |
| # Recaptioned LAION |
|
|
| A subset of LAION recaptioned with **Gemma-3-27B-IT**. Each image is paired |
| with a single descriptive sentence (~20–30 words) generated by Gemma using |
| the prompt: |
|
|
| > *Describe this image in 20-30 words. Do not include any preamble or |
| > introduction, just the description.* |
|
|
| This dataset is used by the **PuzzleBench** project for vision/text |
| alignment training. |
|
|
| ## Stats |
|
|
| | | | |
| |---|---:| |
| | Samples | 508,025 | |
| | Image resolution | up to 1500×1500 (varies) | |
| | Format | JPEG | |
| | Caption type | Gemma-3-27B recaption (declarative, ~130 chars avg) | |
| | Shards | 30 webdataset `.tar` files (~2.4 GB each) | |
| | Total size | ~72 GB | |
|
|
| ## Format: WebDataset |
|
|
| The dataset ships as **30 webdataset-compatible tar shards**, the de facto |
| standard for large image-text corpora (used by LAION-400M, LAION-5B, CLIP |
| training pipelines, etc.). Each shard is a flat tar containing: |
|
|
| ``` |
| shards-00000.tar |
| ├── 000000.jpg # raw image bytes |
| ├── 000000.txt # caption (utf-8) |
| ├── 000001.jpg |
| ├── 000001.txt |
| └── ... |
| ``` |
|
|
| A `metadata.csv` with the same `image_id, url, caption` is also shipped |
| alongside for non-webdataset consumers. |
|
|
| ## Usage |
|
|
| ### With `webdataset` (streaming, no extraction needed) |
|
|
| ```python |
| import webdataset as wds |
| |
| url = "https://huggingface.co/datasets/PuzzleBench/Recaptioned_LAION/resolve/main/shards-{00000..00029}.tar" |
| ds = ( |
| wds.WebDataset(url) |
| .decode("pil") |
| .to_tuple("jpg", "txt") |
| ) |
| |
| for image, caption in ds: |
| # image: PIL.Image |
| # caption: str |
| ... |
| ``` |
|
|
| ### With `huggingface_hub` snapshot + local extraction |
| |
| ```python |
| from huggingface_hub import snapshot_download |
| |
| snapshot_download( |
| repo_id="PuzzleBench/Recaptioned_LAION", |
| repo_type="dataset", |
| local_dir="recaptioned_laion", |
| ) |
| ``` |
| |
| ### With `pandas` (captions only, no images) |
|
|
| ```python |
| import pandas as pd |
| df = pd.read_csv("hf://datasets/PuzzleBench/Recaptioned_LAION/metadata.csv") |
| # columns: image_id, url, caption |
| ``` |
|
|
| ## License |
|
|
| This dataset of recaptions is released under **CC BY 4.0**. The underlying |
| images are sourced from public LAION URLs and retain their original |
| licensing — please consult the LAION usage terms when redistributing or |
| training on the image content. |
|
|
| ## Citation |
|
|
| If you use this dataset, please cite the TDDN paper (as mentioned below), cite/credit |
| the LAION-5B [paper](https://arxiv.org/abs/2210.08402) and [corpus](https://laion.ai/blog/laion-aesthetics/), and [google/gemma-3-27b-it](https://huggingface.co/google/gemma-3-27b-it) model. |
|
|
| ## Citation |
|
|
| ```bibtex |
| @article{patnala2026tddn, |
| title={{TDDN}: {T}ext-aligned {D}iffused {D}INO {N}etwork for Puzzle Understanding}, |
| author={Harsha Patnala and Debopriyo Banerjee and Ayush Sunil Munot and Somak Aditya}, |
| year={2026}, |
| journal={arXiv:2609.07937} |
| eprint={2609.07937}, |
| archivePrefix={arXiv}, |
| primaryClass={cs.CV}, |
| url={https://arxiv.org/abs/2609.07937}, |
| } |
| |
| |
