TriviaQA: question answering with evidence
650K+ question-answer-evidence triples for English reading comprehension and QA.
- Question answering
- English
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
650K+ question-answer-evidence triples for English reading comprehension and QA.
1,300 final attempts from 13 agent systems on a parametric 3D modeling benchmark in FreeCAD, complete with logs and results.
52k instructions and demonstrations from text-davinci-003 for instruction-tuning language models.
306k tasks and millions of CoT/TIR solutions for training LLM mathematical reasoning.
3D photon waveforms from NEON LiDAR data for geospatial quality improvement and resolution tasks.
Benchmark for multimodal understanding and reasoning featuring questions with images across multiple disciplines.
Over 5,000 hours of driving across 30 German cities with 6-camera video, telemetry, and environment labels.
Multilingual audio and text for speech generation: 101k hours original and over 200k in Emilia-Large.
5.5K data analysis tasks: table, question, and ground truth answer.
Multimodal 3D and video dataset from Seoul National University for learning dexterous grasping.
Tens of millions of molecular structure records for training protein design models.
Large corpus of aligned parliamentary speech in 22 European languages for ASR and TTS.
Math question-answer pairs augmented from GSM8K and MATH train sets for reasoning training.
An English set of 1k–10k examples for evaluating models on video question-answering tasks.
200k filtered English dialogues used to train Zephyr-7B-β.
Teleoperation recordings of a robot with two five-fingered hands for training robotics models.
A collection of images and text labels of road scenes in clear weather, fog, rain, and snow for computer vision systems.
113k Wikipedia question-answer pairs with supporting facts for multi-document answering.
The largest open text corpus from PDFs: 3 trillion tokens, 475 million documents, and 1,733 languages.
Tabular and text dataset of human gene expression profiles during respiratory viral immunization.
50,137 tasks from 128 GitHub repositories for training code repair models.
Continuous WebSocket message stream of Polymarket 5-minute markets and Chainlink prices.
1 to 10 billion source code snippets across 16 programming languages for training code models.
3D CT abdominal benchmark for Visual Question Answering with 438K QA pairs and clinical subtypes.