Prophet's Mosque Library: 48k Arabic Books
Digitized Arabic Islamic books from the Prophet's Mosque Library for OCR and language models.
- Image → text
- Arabic
Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.
Answered by Gemini 4 · the chat is kept only in this tab
Digitized Arabic Islamic books from the Prophet's Mosque Library for OCR and language models.
Synchronized camera, LiDAR, and step-by-step reasoning data for training autonomous driving systems.
24.3 million interleaved image-text examples in English and Chinese for training multimodal models.
Over 10,000 hours of real-world recordings of craftsmanship, cooking, and construction work with IMU sensor data.
A large English WorldBench benchmark for evaluating physical understanding in generative and vision models.
Over 10 thousand Arabic books in PDF, TXT, and DOCX formats, recognized via Google Document AI.
English MINT-1T HTML corpus: part of an open 1 trillion token and 3.4 billion image multimodal dataset.
64 thousand real image-question-answer pairs from StackExchange for VQA tasks with verified answers.
Corpus of over 4.6 thousand Arabic books (over 5 million pages) with extracted text.
A set of CT scans with paired radiology reports for training and evaluating 3D VLMs.
A three-configuration set featuring CadQuery 3D code, part renders, editing tasks, and CAD QA.
Generated images, genomes, and VLM agent reasoning logs from the Picbreeder study.
Huge raw pool of nearly 2 billion image-text pairs in six languages for curating data for VLMs.
Video and text pairs for training and evaluating video question-answering models.
A huge corpus of multimodal instructions (image + text) for full-parameter SFT of vision-language models.
Brain and spine MRIs with text conclusions for multimodal medical models.
Benchmark for multimodal understanding and reasoning featuring questions with images across multiple disciplines.
Data for proactive robot manipulation in an omnimodal context.
Multimodal dataset with images and English/Chinese questions for Chinese porcelain appraisal.
An English set of 1k–10k examples for evaluating models on video question-answering tasks.
A collection of videos and texts for video-text-to-text tasks, currently undergoing cleaning.
3D CT abdominal benchmark for Visual Question Answering with 438K QA pairs and clinical subtypes.
Cleaned mixture of 103.9 million image-text pairs in 6 languages in WebDataset format for VLM pre-training.
A lightweight version of the Unsplash dataset with texts and metadata for 25,000 English photographs.