datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
WorldSpeech
WorldSpeech
A multilingual ASR dataset containing over 65k hours of human transcribed speech across 127 language-region variants, drawn from national parliaments, public broadcasters, public-domain audiobooks, and international institutions. Rows consist of 24 kHz speech utterances paired with a human-provided transcript, an aligned ASR transcript, character error rate (CER) between the two, a WADA-SNR estimate, and four DNSMOS-P.835 quality scores.
Dataset Overview… See the full description on the dataset page: https://huggingface.co/datasets/disco-eth/WorldSpeech.EuroSpeech
EuroSpeech Dataset
Dataset Description
EuroSpeech is a large-scale multilingual speech corpus containing high-quality aligned parliamentary speech across 22 European languages. The dataset was constructed by processing parliamentary proceedings using a robust alignment pipeline that handles diverse audio formats and non-verbatim transcripts. More information can be found in the paper.
This dataset is 16 kHz, the 24 kHz version of EuroSpeech can be found at… See the full description on the dataset page: https://huggingface.co/datasets/disco-eth/EuroSpeech.food101
Dataset Card for Food-101
Dataset Summary
This dataset consists of 101 food categories, with 101'000 images. For each class, 250 manually reviewed test images are provided as well as 750 training images. On purpose, the training images were not cleaned, and thus still contain some amount of noise. This comes mostly in the form of intense colors and sometimes wrong labels. All images were rescaled to have a maximum side length of 512 pixels.
Supported Tasks and… See the full description on the dataset page: https://huggingface.co/datasets/ethz/food101.eurospeech-raw-filesEgoPressureEgoPressure: A Dataset for Hand Pressure and Pose Estimation in Egocentric Vision
CVPR 2025 (Highlight)
Yiming Zhao1*, Taein Kwon1*, Paul Streli1*, Marc Pollefeys1,2, Christian Holz1
1 ETH Zürich
2 Microsoft
* Equal contribution
A dataset for hand pressure and pose estimation in egocentric vision.
Participants perform touch gestures on a
pressure-sensing touchpad (Sensel Morph) while being recorded by a
head-mounted egocentric camera and seven static Azure Kinects — providing
synchronized… See the full description on the dataset page: https://huggingface.co/datasets/eth-siplab/EgoPressure.mv3dpt-datasets
Multi-View 3D Point Tracking Datasets
This repository hosts the training and evaluation datasets associated with the paper Multi-View 3D Point Tracking.
Project Page: https://ethz-vlg.github.io/mvtracker/
Code/Github Repository: https://github.com/ethz-vlg/mvtracker
Abstract
We introduce the first data-driven multi-view 3D point tracker, designed to track arbitrary points in dynamic scenes using multiple camera views. Unlike existing monocular trackers, which struggle… See the full description on the dataset page: https://huggingface.co/datasets/ethz-vlg/mv3dpt-datasets.tsnvbsdnnist-publications-raw
NIST Publications - Raw PDFs
596 NIST cybersecurity publications in original PDF format - Complete source data for the nist-cybersecurity-training dataset.
Dataset Description
This dataset contains the raw, unprocessed PDF files downloaded from the NIST Computer Security Resource Center (CSRC). These are the exact source documents used to create the NIST cybersecurity training dataset and fine-tune the HackIDLE-NIST-Coder model.
Contents
596 PDF documents (2.0… See the full description on the dataset page: https://huggingface.co/datasets/ethanolivertroy/nist-publications-raw.PanoInfinigen🗃️ PanoInfinigen Dataset
PanoInfinigen is a synthetic dataset of high-resolution panoramic images in ERP, featuring perfectly aligned RGB, Depth, and Surface Normals. This dataset was generated using a modified Infinigen framework to support wide-angle panoramic geometry, plus the iCity procedural city generator for the urban split.
It serves as the primary training data for PaGeR, a single-step diffusion model for zero-shot panoramic depth… See the full description on the dataset page: https://huggingface.co/datasets/prs-eth/PanoInfinigen.hendrycks_ethicsThe ETHICS dataset is a benchmark that spans concepts in justice, well-being,
duties, virtues, and commonsense morality. Models predict widespread moral
judgments about diverse text scenarios. This requires connecting physical and
social world knowledge to value judgements, a capability that may enable us
to steer chatbot outputs or eventually regularize open-ended reinforcement
learning agents.ethicsA benchmark that spans concepts in justice, well-being, duties, virtues, and commonsense morality.stable-bias-generationsipfs_ethiopia_laws
Laws of Ethiopia
Research snapshot of official legislation collected from Federal Negarit Gazette.
Not legal advice. Official gazettes / government portals prevail over this corpus.
Snapshot
Field
Value
Snapshot date
2026-09-22
Coverage
catalog-backed incomplete
Source
Federal Negarit Gazette
Collector
scrapers/collect_et.py
Laws / instruments
5164
Articles
19221
Language
am
Jurisdiction
Ethiopia
License
et-negarit… See the full description on the dataset page: https://huggingface.co/datasets/endomorphosis/ipfs_ethiopia_laws.AIMEfrom datasets import load_dataset
dataset = load_dataset('disco-eth/AIME')
AIME: AI Music Evaluation Dataset
The AIME dataset contains 6,000 audio tracks generated by 12 music generation models in addition to 500 tracks from MTG-Jamendo.
The prompts used to generate music are combinations of representative and diverse tags from the MTG-Jamendo dataset.
The AIME dataset consists of two subsets. The AIME audio dataset and the AIME survey dataset.
The dataset contains the following… See the full description on the dataset page: https://huggingface.co/datasets/disco-eth/AIME.poly-eth-orderbookMem-Gallery
📖 Overview
Mem-Gallery is a comprehensive benchmark dataset designed to evaluate multimodal long-term memory capabilities of MLLM agents across multi-session conversations. The dataset features realistic, persona-driven dialogues spanning 20 scenarios, each enriched with contextual images to test memory retention, recall, and reasoning over extended interactions.
🎯 Key Features
Diverse Scenarios: Covering topics from AI & Robotics to Daily Life… See the full description on the dataset page: https://huggingface.co/datasets/Ethan-Bei/Mem-Gallery.Transactions-Ethereum-Cryptocurrency-Data
Transactions-Ethereum-Cryptocurrency-Data
Hive-partitioned Parquet export of BlockDB transactions (Ethereum).
Load
from datasets import load_dataset
ds = load_dataset("RBudzynski/Transactions-Ethereum-Cryptocurrency-Data", split="train")
Files live under data/year=YYYY/month=MM/part-NNNN.parquet.
Range: 2015-08 .. 2026-06 (UTC calendar months).
Schema
column
type
block_timestamp
timestamp
block_number
int64
tx_index
int32
tx_hash… See the full description on the dataset page: https://huggingface.co/datasets/RBudzynski/Transactions-Ethereum-Cryptocurrency-Data.nist-cybersecurity-training
NIST Cybersecurity Training Dataset v1.1
The largest open-source NIST cybersecurity training dataset for fine-tuning LLMs
Version 1.1 Highlights
What's New in v1.1:
✅ Added CSWP (Cybersecurity White Papers) series - 23 new documents
✅ Fixed 6,150 broken DOI links via format normalization
✅ Removed 202 malformed DOIs (double URL prefixes)
✅ Validated and fixed 124,946 total links
✅ Cataloged 72,698 broken links for future recovery
✅ 0 broken link markers remaining in… See the full description on the dataset page: https://huggingface.co/datasets/ethanolivertroy/nist-cybersecurity-training.ethereum-gas-estimator-accuracy
Ethereum gas estimator comparisons
Ethereum mainnet gas suggestions from public RPC providers, recorded alongside provider-reported head blocks and subsequent block-fee statistics. The panel supports analysis of differences in gas recommendations and their relationship to observed block conditions.
Contents
Table
Record
ethereum_gas_recommendations
A provider's gas-price and priority-fee suggestions, with its reported head block
ethereum_block_fees
A… See the full description on the dataset page: https://huggingface.co/datasets/dataforge-labs/ethereum-gas-estimator-accuracy.AGBD_rawEuroSpeech-24kHz
EuroSpeech 24 kHz Dataset
Dataset Description
EuroSpeech is a large-scale multilingual speech corpus containing high-quality aligned parliamentary speech across 22 European languages. The dataset was constructed by processing parliamentary proceedings using a robust alignment pipeline that handles diverse audio formats and non-verbatim transcripts. More information can be found in the paper.
Dataset Summary
Languages: 22 European languages (see detailed… See the full description on the dataset page: https://huggingface.co/datasets/disco-eth/EuroSpeech-24kHz.ethereum-attestation-pool
Ethereum attestation pool
Slot-level observations comparing attestations visible in a consensus client's pending pool with attestations subsequently included in blocks. The panel supports analysis of local pool coverage and inclusion counts.
Contents
Table
Record
ethereum_attestation_observations
Attester-slot counts observed pending and included, with their difference and collection coverage
Using the data
attesters_seen_in_pool… See the full description on the dataset page: https://huggingface.co/datasets/dataforge-labs/ethereum-attestation-pool.GlobalDISCO
GlobalDISCO
GlobalDISCO is a large-scale dataset consisting of 73k music tracks generated by state-of-the-art commercial generative music models, along with paired links to 93k reference tracks in LAION-DISCO-12M. The dataset spans 147 languages and includes musical style prompts extracted from MusicBrainz and Wikipedia. The dataset is globally balanced, representing musical styles from artists across 79 countries and five continents. It is aimed to support the research community in… See the full description on the dataset page: https://huggingface.co/datasets/disco-eth/GlobalDISCO.ethereum-mempool
Ethereum mempool observations
Ethereum pending-transaction activity, lifecycle observations and comparisons between provider views. The continuing panels contain minute-level summaries, transactions classified as dropped and differences between observed pending sets.
Contents
Table
Record
ethereum_mempool_minute_summary
A minute of pending-transaction activity
ethereum_dropped_transactions
A transaction classified as dropped during observation… See the full description on the dataset page: https://huggingface.co/datasets/dataforge-labs/ethereum-mempool.Logs-Ethereum-Cryptocurrency-Data
Logs-Ethereum-Cryptocurrency-Data
Hive-partitioned Parquet export of BlockDB logs (Ethereum).
Load
from datasets import load_dataset
ds = load_dataset("RBudzynski/Logs-Ethereum-Cryptocurrency-Data", split="train")
Files live under data/year=YYYY/month=MM/part-NNNN.parquet.
Range: 2015-08 .. 2026-06 (UTC calendar months).
Schema
column
type
block_timestamp
timestamp
block_number
int64
tx_index
int32
log_index
int32… See the full description on the dataset page: https://huggingface.co/datasets/RBudzynski/Logs-Ethereum-Cryptocurrency-Data.gyrollava-665k
LLaVA-v1.5 Mix665K — Arrow (images embedded)
The LLaVA-v1.5 visual instruction-tuning mixture (llava_v1_5_mix665k) converted to a 🤗 datasets
Arrow dataset with image bytes embedded. 665,298 examples
(624,610 image–text + 40,688 text-only).
⚠️ This repo is a raw save_to_disk Arrow snapshot. The dataset viewer and
load_dataset() do not work here — load it with load_from_disk as shown below.
Loading
from huggingface_hub import snapshot_download
from datasets… See the full description on the dataset page: https://huggingface.co/datasets/Ethlake/llava-665k.ethosETHOS: onlinE haTe speecH detectiOn dataSet. This repository contains a dataset for hate speech
detection on social media platforms, called Ethos. There are two variations of the dataset:
Ethos_Dataset_Binary: contains 998 comments in the dataset alongside with a label
about hate speech presence or absence. 565 of them do not contain hate speech,
while the rest of them, 433, contain.
Ethos_Dataset_Multi_Label: which contains 8 labels for the 433 comments with hate speech content.
These labels are violence (if it incites (1) or not (0) violence), directed_vs_general (if it is
directed to a person (1) or a group (0)), and 6 labels about the category of hate speech like,
gender, race, national_origin, disability, religion and sexual_orientation.