datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Tunisian-Proverbs-with-Image-Associations-A-Cultural-and-Linguistic-DatasetTunisian Proverbs with Image Associations: A Cultural and Linguistic Dataset
Description
This dataset explores the rich oral tradition of Tunisian proverbs mapped into text format, pairing each with contextual explanations, English translations both word-to-word and it's equivalent Target Language dynamic, Automated prompt and AI-generated visual interpretations.
It bridges linguistic, cultural, and visual modalities making it valuable for tasks in cross-cultural NLP, generative… See the full description on the dataset page: https://huggingface.co/datasets/HabibaAbderrahim/Tunisian-Proverbs-with-Image-Associations-A-Cultural-and-Linguistic-Dataset.OCRBench_v2lingbot-depth-subset
Dataset Card for lingbot-depth-subset
This is a FiftyOne dataset with 13,149 samples
(10,207 groups) spanning 3 sub-collections (RobbyReal, RobbyVla, RobbySim).
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import load_from_hub
# Load the dataset
# Note: other available arguments include 'max_samples', etc
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/lingbot-depth-subset.3DHarnessBench
Probing Agentic 3D-to-Code Capabilities of Frontier Vision-Language Models
Project Page
·
GitHub
·
GitHub
Abstract
3DHarnessBench evaluates the agentic capacity of frontier vision-language models (VLMs) to recover 3D geometry as executable Blender Python code from multiple forms of target evidence. Rather than restricting every system to a single fixed input, the benchmark compares four progressively richer harnesses: Single-view, Multi-view, Active… See the full description on the dataset page: https://huggingface.co/datasets/lingada/3DHarnessBench.lingyulingbot-map-demoMOT17
MOT17
MOT17 is a benchmark dataset for single-camera multi-object tracking (MOT), focused primarily on pedestrian tracking in real-world video sequences. This Hugging Face repository provides the MOT17 data in the original MOTChallenge-style structure for research, benchmarking, training, and evaluation of multi-object tracking systems.
MOT17 extends MOT16 with more accurate ground-truth annotations and provides each sequence with three public detection sets:
DPM
Faster R-CNN /… See the full description on the dataset page: https://huggingface.co/datasets/ling1016/MOT17.mqud
MQUD: Multimodal Questions Under Discussion
MQUD contains 1,250 figure-grounded inquisitive questions from scientific papers.
Each example pairs a scientific figure with paper context, a question, an
extractive answer, question type, and author-facing metadata.
Paper: Multimodal QUD: Inquisitive Questions from Scientific Figures.
Files
data/mqud.jsonl: one row per MQUD question.
data/mqud.parquet: the same rows in Parquet format when exported with pandas/pyarrow.… See the full description on the dataset page: https://huggingface.co/datasets/lingchensanwen/mqud.ramanv-image-editing
ramanv-image-editing
Image editing dataset for training FLUX.1-Kontext / InstructPix2Pix style models.
Size
592,141 total editing pairs
Sources: ultraedit
Schema
Each shard tar contains {uid}_src.jpg, {uid}_edit.jpg, {uid}_mask.png (where available).
Metadata per record: instruction, prompt, edit_type, caption_before/after, license, sha256.
Licenses
MagicBrush, InstructPix2Pix, Pico-Banana, HumanEdit: CC-BY-4.0
UltraEdit, AnyEdit… See the full description on the dataset page: https://huggingface.co/datasets/lingamvamshikrishnareddy/ramanv-image-editing.rendered-sts17
Dataset Summary
This dataset is rendered to images from STS-17. We envision the need to assess vision encoders' abilities to understand texts. A natural way will be assessing them with the STS protocols, with texts rendered into images.
Examples of Use
Load Arabic to Arabic dataset:
from datasets import load_dataset
dataset = load_dataset("Pixel-Linguist/rendered-sts17", name="ar-ar", split="test")
Load French to English dataset:
from datasets import load_dataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/Pixel-Linguist/rendered-sts17.coyo-700m
Dataset Card for COYO-700M
Dataset Summary
COYO-700M is a large-scale dataset that contains 747M image-text pairs as well as many other meta-attributes to increase the usability to train various models. Our dataset follows a similar strategy to previous vision-and-language datasets, collecting many informative pairs of alt-text and its associated image in HTML documents. We expect COYO to be used to train popular large-scale foundation models
complementary to other… See the full description on the dataset page: https://huggingface.co/datasets/lingao123/coyo-700m.triveni-raw
📦 Pretraining Corpus
📊 Dataset Overview
This dataset combines data from two major sources—Vaani and Flickr30k—to support multilingual and multimodal model pretraining.
Source
Languages
Samples per Language
Total Samples
Vaani
Hindi, English, Hinglish
30,195
90,585
Flickr30k
Hindi, English, Hinglish
31,014
93,042
Total
—
—
183,627
📁 Dataset Sources
🗣️ Vaani Dataset
License: CC-BY-4.0
Description:
VAANI is an… See the full description on the dataset page: https://huggingface.co/datasets/LingoIITGN/triveni-raw.Robot_physics_interaction_with_Human_linguistic
Dual-Camera Speech and Interaction Dataset
This directory contains the processed output of 50 synchronized recording sessions. Each session combines first-person video, third-person video, human speech, 30 fps image sequences, and a text label derived from the read-aloud script.
For the research context behind the dataset, see PROJECT_README.md.
Directory structure
ACTION_NUMBER/
├── first_person_view.mp4
├── third_person_view.mp4
├── audio.wav
├── text.txt
├──… See the full description on the dataset page: https://huggingface.co/datasets/iciclab/Robot_physics_interaction_with_Human_linguistic.rendered-stsb
Dataset Summary
This dataset is rendered to images from STS-benchmark. We envision the need to assess vision encoders' abilities to understand texts. A natural way will be assessing them with the STS protocols, with texts rendered into images.
Examples of Use
Load English train Dataset:
from datasets import load_dataset
dataset = load_dataset("Pixel-Linguist/rendered-stsb", name="en", split="train")
Load Chinese dev Dataset:
from datasets import load_dataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/Pixel-Linguist/rendered-stsb.SangrahaTox
[!CAUTION]
Important Disclaimer & Ethical Notice
1. Research Use OnlyThis dataset is provided strictly for academic and research purposes. Any commercial use or redistribution is prohibited.
2. Content WarningThis dataset contains sensitive information and graphic visuals. Users are advised that the content may be distressing or inappropriate for certain audiences. Viewer discretion is strongly advised.
3. Limitation of LiabilityThe authors and contributors make no warranties… See the full description on the dataset page: https://huggingface.co/datasets/LingoIITGN/SangrahaTox.Triveni
📦 Pretraining Corpus
📊 Dataset Overview
This dataset combines data from two major sources—Vaani and Flickr30k—to support multilingual and multimodal model pretraining.
Source
Languages
Samples per Language
Total Samples
Vaani
Hindi, English, Hinglish
30,195
90,585
Flickr30k
Hindi, English, Hinglish
31,014
93,042
Total
—
—
183,627
📁 Dataset Sources
🗣️ Vaani Dataset
License: CC-BY-4.0
Description:
VAANI is an… See the full description on the dataset page: https://huggingface.co/datasets/LingoIITGN/Triveni.ramanv-domain-trainingProject-Imaging-X
Project Imaging-X is a strategic initiative to consolidate 1000+ open medical imaging datasets worldwide, breaking down data silos through systematic integration to build the foundational infrastructure for next-generation medical AI models.
Challenge: Medical imaging lacks large-scale unified datasets due to clinical expertise requirements and privacy constraints, limiting the development of powerful medical foundation models.
Solution: We surveyed 1000+… See the full description on the dataset page: https://huggingface.co/datasets/lingcarzy/Project-Imaging-X.lingoqaMultimodalCodingDataset
Dataset Card for Multimodal Coding Dataset (MCD)
The Multimodal Coding Dataset (MCD) is a large-scale dataset for training and evaluating multimodal code generation models.
It contains 598k high-quality samples across HTML generation, chart-to-code, image-augmented QA, and algorithmic problems.
This dataset supports research in unifying vision-language understanding with code generation, as described in our paper "VisCodex: Unified Multimodal Code Generation via Merging… See the full description on the dataset page: https://huggingface.co/datasets/lingjie23/MultimodalCodingDataset.ramanv-image-rawqwen-vl-lingala-dataset-augmented
Augmentation
This dataset derives from dataset-qwen-vl-lingala-qlora-vf (417 train / 50 test) through an augmentation step applied to the training image-text pairs, bringing the training volume to 884 examples.
Augmentation method: the 467 additional training examples compared to the source (417 → 884) are obtained mainly through controlled degradation of the input image — noise, brightness/contrast variation, light blur — rather than through synthetic content generation or text… See the full description on the dataset page: https://huggingface.co/datasets/Congo-digital-service/qwen-vl-lingala-dataset-augmented.dataset-qwen-vl-lingala-qlora-vf
Qwen-VL Lingala OCR Dataset
Description
Image/text pairs for training a Qwen2-VL model to perform OCR on Lingala text, including the two special characters absent from the standard Latin alphabet: ɔ (U+0254) and ɛ (U+025B).
train: original + augmented images (noise, brightness/contrast, light blur), with targeted oversampling of lines containing ɔ/ɛ.
test: original, non-augmented images only, held out before any oversampling to avoid data leakage.… See the full description on the dataset page: https://huggingface.co/datasets/Congo-digital-service/dataset-qwen-vl-lingala-qlora-vf.ling-3.0-tiny-atlas
ling-3.0-tiny-atlas
multi-lingual-handwritten-ocr-datasetDocV3R-BenchLingT2I
LingT2I (TRIG-Multilingual)
LingT2I (TRIG-Multilingual) is a multilingual image-generation benchmark with two subsets:
content_generation: multilingual prompts for evaluating generated image content across dimensions.
text_rendering: multilingual prompts for evaluating rendered text in generated images.
The original JSON files are preserved in this repository for traceability:
raw/text-to-image-multilingual.json
raw/trig_multilingual_tr.json
raw/coarse_mask.zip
The public… See the full description on the dataset page: https://huggingface.co/datasets/RISys-Lab/LingT2I.StableDiffusion-v1-5-Regularization-ImagesA collection of regularization / class instance datasets for the Stable Diffusion v1-5 model to use for DreamBooth prior preservation loss training. Files labeled with "mse vae" used the stabilityai/sd-vae-ft-mse VAE. For ease of use, datasets are stored as zip files containing 512x512 PNG images. The number of images in each zip file is specified at the end of the filename.
There is currently a bug where HuggingFace is incorrectly reporting that the datasets are pickled. They are not picked… See the full description on the dataset page: https://huggingface.co/datasets/lingcarzy/StableDiffusion-v1-5-Regularization-Images.dreambooth
Dataset Card for "dreambooth"
Dataset of the Google paper DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
The dataset includes 30 subjects of 15 different classes. 9 out of these subjects are live subjects (dogs and cats) and 21 are objects. The dataset contains a variable number of images per subject (4-6). Images of the subjects are usually captured in different conditions, environments and under different angles.
We include a file… See the full description on the dataset page: https://huggingface.co/datasets/lingcarzy/dreambooth.CHUBS
CHUBS: A Large-Scale Dataset of Chu Bamboo Slip Script
Code | Paper (upcoming)
Introduction
This is a large-scale dataset of Chu bamboo slip (CBS, Chinese: 楚简, chujian) script, an ancient Chinese script used during the Spring and Autumn period over 2,000 years ago. This dataset consists of two parts:
The main dataset where each example is an image and the corresponding text label. This part is contained in the glyphs.zip ZIP file.
A character detection dataset… See the full description on the dataset page: https://huggingface.co/datasets/LINGJIAN5028/CHUBS.