Loading...
Loading...
News corpora, social media analysis, movie/music metadata, sports data, cultural datasets, misinformation
11,206 datasets
This Tatar language resource directory from neurotatarlar aggregates links to translation tools and linguistic datasets as of January 2026. It functions as a curated hub for the Tatarstan region, organizing cultural and technical projects into a single repository.
8,248 anime entries document titles and related metadata from 1917 to 2025. This sample dataset was curated by DivyanshuSingh96 and published on Hugging Face in November 2025. It serves as a preview for a larger recommendation system project.
A cleaned version of the RefCOCO (UNC) validation split for referring expression segmentation. The dataset replaces original instance masks to address poor mask quality and harmful referring expressions, which made the original benchmark an imprecise measure for modern models. It was created by author 'moondream' and last updated on 2025-11-17.
Featuring Twitter activity data for 45 Spanish food industry brands. It includes variables such as Date, Handle, Brand Name, Text, Retweet count, and Favorite count, linked to a research project on brand content dissemination.
InternRobotics released MotionMillion in 2025, providing over 2,000 hours of human motion captured from web-scale video data. The collection spans diverse activities including Fitness (26.4%), Martial Arts (23.7%), and Dance (14.9%), featuring more than 20 detailed semantic annotations per sequence.
Two annotated files contain 1,939 total tokens of Spanish vocatives for apparent-time and micro-diachronic analysis. The first file has 1,107 tokens from the CORMA corpus recorded between 2016 and 2019, annotated for form and speaker generation. The second file contains 832 tokens from corpora spanning the 1980s to the 2000s, annotated for form, corpus, and decade. The data was created by author De Latte, Fien and is hosted by DataverseNO.
People's Daily News is a dataset of articles from the Chinese newspaper People's Daily, covering a period from 1946 to 2024. The dataset is part of the CialloCorpus collection and was uploaded by shenxs to Hugging Face in December 2025.
Dolmino Mix 1124 is a mixture of high-quality data used for the second stage of OLMo2 training. The dataset aggregates text from multiple sources, including 752 billion tokens from DCLM web pages, 58.6 billion tokens from Pes2o STEM papers, and 1.26 billion tokens from StackExchange code text. It was created by AllenAI and last updated on October 29, 2025.
Image assets and metadata categorized into single-food crops, landmarks, and street scenes for benchmarking Large Vision-Language Models. The collection focuses on culture mixing scenarios where multiple cultural identifiers co-exist within a single visual frame as defined in the World in a Frame benchmark.
Survey data tracks the proportion of adults in California ever diagnosed with a depressive disorder. The information originates from the annual California Behavioral Risk Factor Surveillance Survey (BRFSS), conducted via telephone by California State University, Sacramento under contract from the state's Department of Public Health. The data uses state-specific population denominators from the California Department of Finance.
82 video clips of five exercise types, totaling 8,029 frames, annotated for temporal action classification and form analysis. The dataset, created by AvihaiNaam, is split into 54 training and 28 validation videos. It was last updated on Hugging Face in November 2025.
A dataset for training AI models on spatial understanding and block construction tasks. The dataset is provided as a compressed archive named 'semantic_blocks_part2.tar.gz' with a size of 0.81 GB. It was uploaded by author 'yinbq' and last updated on November 4, 2025.
MM-Health is a large-scale multimodal dataset containing 34,746 news articles with both textual and visual information. It is designed for benchmarking the detection of health misinformation, including content generated by both humans and AI. The dataset was created by author zzha6204 and was last updated on huggingface in November 2025.
MohammadGholizadeh published a dataset titled 'Bplus Podcast Persian' on HuggingFace. The dataset likely contains audio files or transcripts from Persian-language podcasts. Its content and scale require verification after download.
An 870-hour voice conversion dataset contains spontaneous, in-the-wild podcast speech. It was curated by JHU-SmileLab in collaboration with the MSP Lab at CMU LTI for training and evaluating emotion-aware voice conversion systems.
Movarekhpodcast Dataset Persian is a collection of audio and likely associated text from Persian-language podcasts. It was authored by MohammadGholizadeh and published on the Hugging Face platform, with a last recorded update on December 21, 2025. The dataset's specific content, scale, and structure require verification after download.
Turkish Podcast Merge V2 is a dataset of Turkish-language podcast audio and likely associated metadata, hosted on HuggingFace. The dataset was uploaded by the author Aynursusuz and was last updated on December 21, 2025. Its specific size, format, and content details are not provided in the available metadata.
Vietnamese Cultural VQA Dataset is a multimodal collection focusing on Vietnamese cultural heritage, created by Dangindev and last updated on November 13, 2025. It contains 28,505 images across 12 cultural categories paired with 119,012 question-answer pairs in Vietnamese and English. The dataset covers diverse aspects including architecture, cuisine, traditional clothing, landscapes, festivals, and folk culture.
A dataset titled 'Fusion Dataset Gemini 3 Pro Preview Sft' was published on huggingface by DesmondZhengtingHe. The dataset was last updated on 2025-12-24 02:44:27. Its specific content, size, and structure are not described in the available metadata.
Sports-QA is a benchmark dataset proposed in the paper "Sports-QA: A Large-Scale Video Question Answering Benchmark for Complex and Professional Sports". The dataset contains videos and is hosted on HuggingFace by author HopLeeTop, with a last recorded update in December 2025. Further details are referenced to be available on a GitHub page.