To search datasets programmatically: GET https://api.databazaar.io/datasets?query=your-search
Full API docs: https://api.databazaar.io/llms.txt
Agent discovery: https://databazaar.io/.well-known/agent.json
Browse Data
25–48 of 239Hepatitis Prognosis — 1M Synthetic (BNG)
A 1,000,000-row synthetic expansion (generated via a Bayesian Network) of the UCI Hepatitis dataset, predicting patient survival from clinical and laboratory attributes.
German Credit Risk — 1M Synthetic (BNG)
A 1,000,000-row synthetic expansion (generated via a Bayesian Network) of the German Credit dataset, profiling loan applicants by financial and personal attributes with a good/bad credit-risk label.
Space Shuttle Autolanding Control (NASA / UCI)
A small NASA decision dataset specifying the conditions — stability, error, sign, wind, magnitude, and visibility — under which a Space Shuttle should land automatically versus manually. A classic UCI benchmark.
Page Blocks Classification — 295K Synthetic (BNG)
A 295,245-row synthetic expansion (generated via a Bayesian Network) of the classic Page Blocks dataset, classifying blocks in scanned document layouts by geometric and pixel-density features.
Heart Disease (Statlog) — 1M Synthetic (BNG)
A 1,000,000-row synthetic expansion (generated via a Bayesian Network) of the Statlog Heart dataset, predicting heart-disease presence from 13 clinical attributes.
Automobile Price Prediction (UCI)
159 automobiles described by 15 engineering and design attributes — wheelbase, dimensions, engine size, horsepower, fuel economy, and more — with price as the prediction target. A classic UCI regression benchmark.
Heart Disease — Cholesterol Prediction (UCI)
A 303-patient variant of the UCI heart-disease data with serum cholesterol as the prediction target alongside 13 clinical attributes. Used as a numeric (regression) prediction benchmark.
E. coli Promoter Gene Sequences (UCI)
106 E. coli DNA sequences labeled as promoter or non-promoter, each described by 57 sequential nucleotide positions. A classic UCI molecular-biology classification benchmark for sequence analysis.
SEA Concept-Drift Stream — 1M Synthetic Samples
A 1,000,000-row synthetic data stream from the classic SEA generator, with three numeric attributes and a binary class label. A standard benchmark for evaluating concept-drift detection and streaming classifiers.
Statlog Heart Disease (UCI)
The Statlog Heart dataset — 270 patients described by 13 clinical attributes (age, sex, chest-pain type, blood pressure, cholesterol, ECG results, and more) with a heart-disease presence label. A classic UCI classification benchmark.
OpenPII — 1.4M Multilingual PII Masking Examples
1,428,143 synthetic text examples with fine-grained PII annotations spanning 23 European languages. Each example pairs source and masked text with privacy masks and token-level labels, designed for training privacy-redaction and anonymization models.
OPUS-100 — English-Centric Parallel Corpus (100 Languages)
A large English-centric multilingual translation corpus covering 100 languages, where every training pair includes English on the source or target side. A standard resource for machine-translation research and multilingual model training.
Conceptual Captions — 5.3M Image–Caption Pairs
Millions of web images paired with natural-language captions, harvested and filtered from alt-text descriptions. A large-scale resource for training and evaluating image-captioning and vision-language models.
Nemotron Cascade-2 — SFT Training Data (Math, Code & Chat)
~1.94M supervised fine-tuning examples used to train NVIDIA's Nemotron-Cascade-2 models, spanning math, code, and conversational prompts paired with model-generated responses and source/domain labels.
WideSearch — Agentic Broad Information-Seeking Benchmark
A 200-instance benchmark for evaluating LLM-driven agents on broad, wide-coverage information-seeking tasks. Each instance provides a query, structured evaluation criteria, and language metadata for measuring agent search performance.
Emotion — English Tweets Labeled by Emotion
English Twitter messages labeled with six basic emotions — anger, fear, joy, love, sadness, and surprise. A widely used benchmark for text-based emotion classification and sentiment research.
OpenMath GSM8K (Masked) — Grade-School Math Solutions
A masked version of the GSM8K grade-school math word problems (7,473 items), pairing each question with masked and full reference solutions plus the expected answer. Designed to support consistent synthetic generation of additional solutions.
MiniPile — 1M-Document Deduplicated Text Corpus
A compact 6 GB subset of the deduplicated Pile corpus (~1,010,500 documents), curated for data-efficient language-model pretraining and experimentation without the storage cost of the full Pile.
Vessel Detection — Labeled Satellite Image Patches
1,920 validated satellite image patches labeled for maritime vessel detection, exported from a boat-detection review workflow. Each patch ships with Hugging Face image-folder-compatible metadata for object-detection and classification training.
PII Masking Dataset — 300K Labeled Text Samples
225,000+ text samples annotated for personally identifiable information (PII) detection and masking, built to train and evaluate models that redact sensitive data from text. Each record pairs source and target text with privacy masks, span labels, and token-level annotations across multiple languages.
Stack Overflow Posts — 58M Questions & Answers (Markdown)
Every Stack Overflow post submitted before June 14, 2023 — roughly 58 million questions and answers (~35 GB) formatted as Markdown. Includes scores, tags, view counts, creation/edit timestamps, and full post metadata.
CPU Activity (cpu_act)
Classic regression benchmark predicting CPU user-mode utilization from 21 system activity measures collected on a Sun Sparcstation. 8,192 rows, widely used in ML evaluation.
Yeast Protein Localization (UCI/OpenML)
Classic multi-class classification benchmark predicting cellular localization sites of yeast proteins from 8 sequence-derived numeric features. 1,484 instances, 10 classes.
Waveform-5000 (UCI / OpenML)
Classic 3-class synthetic waveform classification benchmark with 5,000 instances and 40 numeric attributes (21 informative + 19 noise). Originally from Breiman et al. 1984, distributed via UCI and OpenML.