Cold-start SFT for image reasoning
Cold-start dataset for SFT models learning multi-turn reasoning with images.
Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,797 in the catalog so far, with new ones added every hour.
Answered by Gemini 4 · the chat is kept only in this tab
Cold-start dataset for SFT models learning multi-turn reasoning with images.
Point-based annotations of landscape changes using Sentinel-2 time-series imagery from AllenAI.
Acapella vocal audio recordings with detailed expert evaluations across nine musical criteria.
Over 19,000 videos of Chinese announcers for training and evaluating portrait video generation models with gestures.
An initiative to unify over 1,000 open medical imaging and document datasets.
A massive corpus of 57.3M short and long English captions for 10M Flickr images from various VLM models.
Color image dataset of industrial fabrics for unsupervised defect detection.
40 generated 3D assets of American houses, offices, and industrial buildings with source renders.
623,957 discovered RSS/Atom feeds from 1.75M Hacker News websites in parquet format.
A benchmark for evaluating goal-directed manipulations in video generation using egocentric frames.
A collection of instruction-based code generation examples for Three.js and Blender.
Synthetic images of Japanese ink paintings for studying diffusion models.
Class image set for prior preservation loss in DreamBooth when fine-tuning Stable Diffusion v1-5.
Invoice and receipt images with text for OCR; limited description in the card.
Corporate document pages with sensitive data markup for tokenization and detection.
An SQLite database of Danbooru2023 tags and posts with a latest ID of 7,866,491 for generative models.
20,000 synthetic landscape images in pixel art style with text prompts.
Corpus of 20,600 pairs of poster images and corresponding SVG files for editable reconstruction.
20k 24x24 pixel art sprites with text captions and semantic taxonomy for image generation.
Manually verified image and transcription pairs of handwritten Vietnamese lines for OCR tasks.
Over 52k cropped text fragments from English scanned receipts for training OCR models.
Documents with text at various angles for OCR and recognition testing.
Around 9,700 full-size webpage screenshots with UI element annotations in YOLO format for interface detection.
A multimodal benchmark of H&E patches and gene expression profiles for image classification tasks.