datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
blog_authorship_corpusNeapolitan-Spoken-Corpus
Neapolitan Spoken Corpus (NSC)
A corpus of read Neapolitan speech for ASR evaluation, with a validated
Neapolitan–Italian lexicon, LOSO fine-tuning splits, trained LoRA adapters,
metric implementations, per-clip results, and error annotations.
This release supersedes the earlier 141-clip single-speaker version of this
repository. The earlier release corresponds to Speaker S1 of the present
corpus; the old audioData/ and transcripts.csv are replaced by
data/audio/ and… See the full description on the dataset page: https://huggingface.co/datasets/anonymous-nsc-author/Neapolitan-Spoken-Corpus.Pashto-Textbooks-PDFs-Corpus
Pashto Textbooks and PDFs Corpus
Languages: psLicense: cc-by-4.0Task categories: text-generation, feature-extractionSize categories: n<1K
Summary
This dataset is part of the ZamAI Pashto data collection. It is intended for text-generation, feature-extraction tasks in Pashto.
How to use
from datasets import load_dataset
dataset = load_dataset("tasal9/Pashto-Textbooks-PDFs-Corpus")
print(dataset)
Configs
default: load with… See the full description on the dataset page: https://huggingface.co/datasets/tasal9/Pashto-Textbooks-PDFs-Corpus.gutenberg-poetry-corpusurban-heat-research-corpus
Urban Heat Research Corpus (UHRC) v1.0
What does the world study, invent and report about urban heat? This dataset puts three records of the same problem side by side:
20,422 research papers on urban heat islands and extreme heat in cities (1990–2025) with the claims their abstracts make,
106,458 news articles about heat (2021–2025) coded for 51 subjects, framings and terms, and
4,123 patent families for heat-mitigation technologies (2006–2024) — plus supplementary tables on the… See the full description on the dataset page: https://huggingface.co/datasets/fliarbi/urban-heat-research-corpus.SpatioTemporal-News-Corpus
Spatiotemporal News Dataset
Overview
This dataset contains approximately 1.2 million English-language news headlines and articles sourced from major outlets in the
United States, United Kingdom, Canada, and Australia.
Each entry is annotated with spatial (country of origin) and temporal (date of publication) contexts, designed for training spatiotemporal-aware sentence embeddings,
specifically our Space-Time-MiniLM-v0 model.
The dataset covers the period from January… See the full description on the dataset page: https://huggingface.co/datasets/Artur-B/SpatioTemporal-News-Corpus.permission-command-corpus
Permission Command Corpus
Three views of a command-safety corpus, for local command-risk classification in
front of an LLM or a tool bridge.
gold: trusted rows, 321 in total across three splits
silver_weak_labels: mined weak-label rows from Sigma, LOLBAS, GTFOBins,
Atomic Red Team and Falco, kept as useful but not promoted to gold
review_queue: unresolved rows that should not be treated as trusted
training data
Read this before training on it
Findings from 30… See the full description on the dataset page: https://huggingface.co/datasets/cowWhySo/permission-command-corpus.blog-authorship-corpuseconomic-telegram-news-corpus-2025
Economic Telegram News Corpus 2025
A corpus of 31,292 Russian-language economic news posts collected from 7 major Telegram channels, spanning January 2024 to September 2025. The dataset supports research on economic narrative detection, topic classification, and information diffusion in social media.
Associated Paper
Going Viral: LLM-Based Modeling of Economic Narratives
Dataset Description
The raw collection contains 123,273 posts. The economic corpus was… See the full description on the dataset page: https://huggingface.co/datasets/bruhwalkk/economic-telegram-news-corpus-2025.rag-qa-logs-corpus-data
🧠📚 RAG QA Logs & Corpus (Synthetic)
🧪 Multi-table synthetic RAG telemetry for quality, hallucinations, latency, and cost
A production-style, privacy-safe synthetic dataset that mimics telemetry exported from a real RAG system — from corpus → chunks → retrieval events → eval runs.
✅ Fully synthetic (no real users / orgs / PII).
⚡ Quick facts
Total rows: 103,255 across 6 linked tables
Labels (in eval_runs): is_correct, hallucination_flag, faithfulness_label… See the full description on the dataset page: https://huggingface.co/datasets/tarekmasryo/rag-qa-logs-corpus-data.indo-bloom-corpus
🇮🇩 Indo-Bloom-AQG: A Unified Framework for Controllable Indonesian AQG
⚠️ RESEARCH ARTIFACT STATUS: SILVER VERSION (Work in Progress)
This dataset serves as the preliminary corpus (Silver Standard) for the ongoing Doctoral Dissertation at Universitas Negeri Malang (UM).
Current State: Unannotated / Pre-validation with Heuristic Bloom Labels
Target Final State: Gold Standard (Expert Validated with Bloom's Taxonomy Labels)
🔒 FROZEN — v0.1 Silver
This version is permanently… See the full description on the dataset page: https://huggingface.co/datasets/Firmansyah-Ibrahim/indo-bloom-corpus.setimes-en-tr-aligned-corpus
SETimes EN-TR — Sentence-Aligned, LLM-Cleaned
A cleaned and re-aligned version of the SETimes English-Turkish parallel corpus. The original SETimes data is paragraph-style — each "pair" can contain a headline, a dateline, several body sentences, and a source citation, all glued together on one line. This version splits everything into proper sentence pairs so each row is one English sentence next to its Turkish translation.
144,064 sentence pairs, split into train (142,064)… See the full description on the dataset page: https://huggingface.co/datasets/atahanuz/setimes-en-tr-aligned-corpus.email-politeness-corpus
Email Politeness Corpus
This dataset accompanies the paper:
A Synthetic Request–Reply Email Corpus Annotated with Document-Level Politeness and Sentence-Level Face Acts
Roshad Alipanah, Valentin Barriere, and Jorge BaierFindings of the Association for Computational Linguistics: EMNLP 2026
The corpus consists of synthetic request–reply emails jointly annotated at two levels:
Sentence level: multi-label Face Act annotations grounded in Brown and Levinson's politeness theory.… See the full description on the dataset page: https://huggingface.co/datasets/ralipanah/email-politeness-corpus.moral-number-corpus
A Perspectivist Corpus of Numbers in Social Judgements
This is the dataset for A Perspectivist Corpus of Numbers in Social Judgements.
We constructed a corpus of moral and social judgements (questions are derived from the Commonsense Norm Bank) that asks people to fill in number ranges that do not change a given judgement.
Our corpus was crowdsourced from 30 annotators and contains 898 statements for a total of 3k annotations.
This work adds to available moral and social judgement… See the full description on the dataset page: https://huggingface.co/datasets/Marlon154/moral-number-corpus.llmstxt-corpus
The llms.txt corpus
Measurement data on the llms.txt convention, collected in one run on
5 August 2026.
llms.txt is a plain-text file at a site's root, proposed as a curated map
telling AI systems what the site contains. This is a measurement of what is
actually being published under that name.
Canonical release: https://doi.org/10.5281/zenodo.22859104
This repository mirrors that deposit. Cite the DOI, which always resolves to the
newest version.
Two observations… See the full description on the dataset page: https://huggingface.co/datasets/sempite/llmstxt-corpus.Saudilang-Code-Switch-Corpus
SCC - Saudilang Code-Switch Corpus
The National Center for Artificial Intelligence at the Saudi Data and Artificial Intelligence Authority (SDAIA), published the "SCC" dataset, which stands for "Saudilang Code-Switch Corpus”.
This dataset contains a transcription of general conversations taken from a YouTube podcast "Thmanyah" that has been transcribed by the National Center for Artificial Intelligence in SDAIA. The data features three episodes covering different domains: investment… See the full description on the dataset page: https://huggingface.co/datasets/SDAIANCAI/Saudilang-Code-Switch-Corpus.brazilian-digital-identity-instagram-corpus
Brazilian Digital Identity on Instagram — Thematic Corpus
Authors: Étore Braga e Santos (Unicamp, ORCID 0009-0000-3502-705X) & Pâmella Fernandes de Sá (USP FEA-RP)License: CC BY 4.0 (dataset) — code at https://github.com/Etore-BeS/brazilian-digital-identity-paper (MIT)Version: 1.0.0 — Collection March 7–11, 2026
Dataset Description
Anonymized corpus of 17,287 Instagram comments on three international celebrity posts with substantial Brazilian engagement, annotated… See the full description on the dataset page: https://huggingface.co/datasets/Etore-BeS/brazilian-digital-identity-instagram-corpus.JParaCrawl-Filtered-English-Japanese-Parallel-Corpus
Introduction
This is a LLM-filtered set of the first 1M rows from ntt's JParaCrawl v3 large English-Japanese parallel corpus.
The original JParaCrawl corpus was put together by automated means - aligning Japanese texts with their apparent English translations that were found in-the-wild, on the internet.
Whilst manually browsing the original data, I noticed that there were obvious quality issues that made me anxious about using the dataset at all. Poorly aligned translations… See the full description on the dataset page: https://huggingface.co/datasets/Verah/JParaCrawl-Filtered-English-Japanese-Parallel-Corpus.appropriateness-corpus
The Appropriateness Corpus
The Appropriateness Corpus is a collection of 2191 arguments annotated for appropriateness and its 14 subdimensions derived in the paper Modeling Appropriate Language in Argumentation published at the ACL2023.
Dataset Description
What does Appropriateness mean?
An argument “has an appropriate style if the used language supports the creation of credibility and emotions as well as if it is proportional to the issue.” Their… See the full description on the dataset page: https://huggingface.co/datasets/timonziegenbein/appropriateness-corpus.EF_Chapters_and_Chunks
Dataset Card for SF Nexus Extracted Features: Chapters and Chunks
Dataset Summary
The SF Nexus Extracted Features Chapters and Chunks dataset contains text and metadata from 403 mid-twentieth century science fiction books, originally digitized from Temple University Libraries' Paskow Science Fiction Collection.
After digitization, the books were cleaned using Abbyy FineReader.
Because this is a collection of copyrighted fiction, the books have been disaggregated.
To… See the full description on the dataset page: https://huggingface.co/datasets/SF-Corpus/EF_Chapters_and_Chunks.RomanUrdu-NLP-Sentiment-Corpus
RomanUrdu-NLP-Sentiment-Corpus
Largest Open-Source Roman Urdu Sentiment Dataset with Slang Robustness
Overview
This repository presents the largest publicly available Roman Urdu sentiment analysis dataset, containing 134,052 labeled text samples collected from chats and social media platforms. The dataset is designed to be:
Robust to slang and informal Roman Urdu
High-quality through LLM-assisted labeling and human validation
Balanced across sentiment classes… See the full description on the dataset page: https://huggingface.co/datasets/Khubaib01/RomanUrdu-NLP-Sentiment-Corpus.CA-EN_Parallel_Corpus
Dataset Card for CA-EN Parallel Corpus
Dataset Description
Dataset Summary
The CA-EN Parallel Corpus is a Catalan-English dataset of parallel sentences created to
support Catalan in NLP tasks, specifically Machine Translation.
Supported Tasks and Leaderboards
The dataset can be used to train Bilingual Machine Translation models between English and Catalan in any direction,
as well as Multilingual Machine Translation models.
Languages
The… See the full description on the dataset page: https://huggingface.co/datasets/projecte-aina/CA-EN_Parallel_Corpus.bengal-dharma-corpus
Bengal Dharma Corpus
Evolution of Bengali Devotional Language: a multi-tradition corpus spanning
Old Bengali, Sanskrit, and modern Bengali across Buddhist, Shakta, and Vaishnava
traditions, 8th to 19th century.
Assembled and curated by Joy Bose (joyboseroy), June 2026.
Code and analysis: https://github.com/joyboseroy/bengal-dharma-corpus
Related dataset: joyboseroy/darshana-graph (arXiv:2606.18222)
What this corpus is
This is a curated collection of 75 texts from… See the full description on the dataset page: https://huggingface.co/datasets/joyboseroy/bengal-dharma-corpus.vietnamese-legal-corpus-20k-rawvqfat-indo-corpus
VQFat Indonesian Corpus - Complete SLM Training Data
Author: Vicky FatrianLicense: C-UDA (Computational Use of Data Agreement)Version: 1.1Date: March 6, 2026
Dataset Description
VQFat Indonesian Corpus adalah koleksi dataset komprehensif berbahasa Indonesia yang dirancang khusus untuk membangun dan melatih Small Language Models (SLM) dari nol (from scratch). Dataset ini menyediakan dua pilar utama dalam pelatihan model bahasa: Pre-training Corpus (untuk pengetahuan umum)… See the full description on the dataset page: https://huggingface.co/datasets/vickyfatrian/vqfat-indo-corpus.Medium-Articles-Corpus
Medium Articles Corpus (10K Sample)
The Medium Articles Corpus is a massive, clean dataset of articles scraped from Medium.com. This sample version contains 10,000 articles + and is designed to showcase the quality and structure of the full corpus for researchers and developers.
This is the subset from the large dataset https://crawlfeeds.com/websites/medium/text_data/medium_articles
Dataset Features
This dataset includes the following key features, provided in a… See the full description on the dataset page: https://huggingface.co/datasets/crawlfeeds/Medium-Articles-Corpus.grite-corpus
grite Coordination-Log Corpus
The coordination-log dataset and pre-registered mining toolkit behind
"Before the Pull Request: Mining Multi-Agent Coordination"
(arXiv:2606.19616).
Autonomous coding agents now open millions of pull requests, yet PR-level
telemetry cannot see how concurrent agents coordinate before the PR. This
corpus captures that pre-PR process — task claims, locks, edits, dependencies —
recorded as grite's append-only
coordination log, plus the failure-mode… See the full description on the dataset page: https://huggingface.co/datasets/dipankarsarkar/grite-corpus.sheria-corpus-v1The Sheria Corpus v1 is a curated collection of Kenyan legal case summaries from both the High Court and the Supreme Court, organized to support legal research, data analysis, and natural language processing applications in the legal domain.
For complete access of the dataset, visit http://datalab.datawiseafrica.com/
Contents:
Case metadata: Includes case number, case name, category (e.g., Judgement, Ruling), and decision date.
Court details: Name, level (High or Supreme Court), location, and… See the full description on the dataset page: https://huggingface.co/datasets/datawise-africa/sheria-corpus-v1.ethical-ai-manifesto-corpus
Dataset Card for OpenEthos Transparency Corpus
📌 Dataset Summary
This dataset contains curated text samples regarding AI governance and ethical manifestos. It is designed to help researchers train models that can identify transparency commitments in technical documentation.
⚖️ Ethical Considerations
As part of the OpenEthos Research Collective, this dataset has been audited for:
Privacy: No personally identifiable information (PII) is included.
Bias: We have… See the full description on the dataset page: https://huggingface.co/datasets/Openethos/ethical-ai-manifesto-corpus.multi-bias-news-corpusMulti-Bias News Corpus (Pilot Phase)
This dataset was created as part of the final-year project
"LLM-Powered Multi-Bias Detection in News Media with Lightweight Human-in-the-Loop Evaluation" (Academic Year 2025–26).
It contains a collection of news articles annotated across multiple types of bias. The aim is to move beyond simple left/center/right classification and capture the different ways bias can appear in news reporting.
Dataset Overview
Total articles: 2023
Type: Multi-label bias… See the full description on the dataset page: https://huggingface.co/datasets/AION26/multi-bias-news-corpus.