datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
web3-trading-analysisThis dataset contains web3-related on-chain and off-chain data, which can be used to build quantitative models.
Survivor
📚 FinePDFs-Edu
350B+ of highly educational tokens from PDFs 📄
What is it?
📚 FinePDFs-Edu dataset consists of 350B+ tokens of educational PDFs filtered from 📄 FinePDFs dataset covering 69 languages.
FinePDFs was created using the formula inspired from FineWeb-Edu, we developed an educational quality classifier using annotations generated by Qwen3-235B-A22B-Instruct-2507 for each of 69 languages present in this dataset.
We then used this classifier to retain only the… See the full description on the dataset page: https://huggingface.co/datasets/Web3Survivor/Survivor.imagesWeb3-Dataset
Web3Coders Smart Contracts Dataset – 2026 Edition
Version 2.0 (August 2026) – The largest curated collection of production-grade smart contracts with integrated security audits, gas profiles, vulnerability annotations, and Halal compliance tags for Shariah‑aware Web3 development.
Dataset Description
A comprehensive, multi‑chain dataset of smart contracts (Solidity, Vyper, Rust, and Move) with:
Full source code, bytecode, and ABI
Security audit outcomes (Slither… See the full description on the dataset page: https://huggingface.co/datasets/KurniaKadir/Web3-Dataset.achilles-web3-trainingvipe-web360
ViPE Dataset Release
This dataset contains the camera pose, depth, and intrinsics estimated using ViPE. For more details of the dataset, please refer to the Github link.
Please consider citing the following paper if you found this dataset helpful:
@article{huang2025vipe,
title={Vipe: Video pose engine for 3d geometric perception},
author={Huang, Jiahui and Zhou, Qunjie and Rabeti, Hesam and Korovko, Aleksandr and Ling, Huan and Ren, Xuanchi and Shen, Tianchang and Gao, Jun and… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/vipe-web360.web3_intents_to_ethereum_transactions
Natural Language Web3 Intents to Ethereum Transactions
This dataset contains instruction-following examples for mapping natural language Web3 intents to executable Ethereum transaction plans. Every transaction is decoded from a successful Ethereum mainnet call (March 2025 – January 2026); every intent is either written by humans or generated by an LLM from the decoded transaction.
It is organized into two JSONL splits:
single_step: one user intent mapped to one on-chain action… See the full description on the dataset page: https://huggingface.co/datasets/Intent2Tx/web3_intents_to_ethereum_transactions.web3_diaryafrofinchain-multilingual-web3
AfroFinChain — Multilingual Web3 & Blockchain Dataset
Multilingual Web3 & blockchain dataset in Yoruba, Hausa, Igbo, and Nigerian Pidgin with 1,451 terminology entries and 1,451 conversational Q&A pairs. Designed for LLM fine-tuning, financial literacy, and conversational AI in low-resource African languages. Uses culturally grounded analogies (e.g., ajo, adashi, isusu) to make DeFi concepts actually understandable.
Built with Adaptive Data by Adaption as part of the Adaption… See the full description on the dataset page: https://huggingface.co/datasets/FirstBML1/afrofinchain-multilingual-web3.web3research-datasetsTIP: The data has not been deduplicated yet, please manually deduplicate the data before using it.
web3-oriented-pretraining-dataStatic knowledge
IqWiki:
iqwiki_articles
Binance Academy
binance_academy_articles
Trading Views Article (Crypto)
trading_view_news
trading_view_ideas
Coinmarket cap (Blog)
cmc_news
Other news
bbc_news
cnn_news
fox_news
nbc_news
Commonsense
Crypto/KOL/Meme/Dump/…..(Grey & dark)
None
Terms
crypto_terms
Web3-findings-dataset
Smart Contract Audit Findings
This is raw, semi-structured data — not a ready-to-train dataset. It still requires
further cleaning and preparation (deduplication, severity/label normalization, filtering
low-quality or malformed entries, etc.) before it should be used to train or fine-tune an AI model.
A collection of 23,625 smart-contract security audit findings (bug reports), each with a
title, description, proof-of-concept code, recommendation, and severity rating.… See the full description on the dataset page: https://huggingface.co/datasets/0xSojalSec/Web3-findings-dataset.smart-contract-intent-vul-dataset
Smart Contract Intent & Vulnerability Dataset
This dataset contains a large-scale collection of smart contracts annotated with malicious intents and security vulnerabilities, designed for research on smart contract security analysis and program understanding.
The dataset includes 90,000+ smart contracts collected from Ethereum and Binance Smart Chain (BSC). All data are distributed as a MySQL database dump (.sql.gz) to support flexible querying and large-scale analysis.… See the full description on the dataset page: https://huggingface.co/datasets/web3se/smart-contract-intent-vul-dataset.KSE-Web3K
KSE-Web3K
This dataset is associated with the paper KSE-Web: An Analysis of Hybrid Retrieval and LLM-Assisted Query Expansion for Low-Resource Khmer Semantic Search.
It contains 3,000 cleaned full-text Khmer web documents, 300 manually reviewed Khmer search queries, and silver relevance labels. The dataset is designed for sparse, dense, hybrid, and LLM-assisted retrieval experiments.
Paper: KSE-Web paper
Code: GitHub repository
The repository provides the following files:… See the full description on the dataset page: https://huggingface.co/datasets/Backkh/KSE-Web3K.indonesia-affordable-housing
Indonesia Affordable Housing Dataset
A comprehensive dataset of affordable housing projects across Indonesia, containing detailed information about residential properties, specifications, locations, pricing, and developer data.
Dataset Creator: web3hungry
Dataset ID: web3hungry/indonesia-affordable-housing
License: CC0 1.0
Dataset Overview
This dataset provides extensive data on housing developments throughout Indonesia, covering both subsidized and commercial… See the full description on the dataset page: https://huggingface.co/datasets/web3hungry/indonesia-affordable-housing.Humanoid-Web3-Command-Dataset-v1
Humanoid-Web3-Command-Dataset-v1
Overview
Humanoid-Web3-Command-Dataset-v1 is a structured dataset designed for training NLP models to classify humanoid robot commands into action intents.
This dataset supports robotics automation and Web3 AI systems.
Dataset Structure
The dataset contains labeled human instructions mapped to predefined action intents.
Labels
move_object
pick_and_place
rotate_object
scan_object
grab_object
release_object… See the full description on the dataset page: https://huggingface.co/datasets/Caplin43/Humanoid-Web3-Command-Dataset-v1.web3defender-crypto-security
Web3Defender Crypto Security Threat Dataset
Educational dataset of Web3 / crypto wallet security threats and defensive guidance,
curated by Web3Defender — an independent defensive knowledge base.
What this is
Instruction-response pairs covering common crypto wallet threats (approval phishing,
honeypot tokens, wallet drainers, seed-phrase scams, address poisoning, fake support)
and how users recognize and defend against them. Defensive education only.… See the full description on the dataset page: https://huggingface.co/datasets/Web-kot/web3defender-crypto-security.web3-knowledge-baseHumanoid-Web3-Advanced-Command-Dataset-v2
Humanoid-Web3-Advanced-Command-Dataset-v2
Overview
Humanoid-Web3-Advanced-Command-Dataset-v2 is a structured NLP dataset created for training humanoid robot intent classification models.
The dataset contains diverse natural language commands mapped into structured action intents for robotics automation and Web3-integrated AI systems.
Task
Text Classification (Intent Detection)
Language
English
Total Samples
Train: 60
Test: 20
Total: 80… See the full description on the dataset page: https://huggingface.co/datasets/Caplin43/Humanoid-Web3-Advanced-Command-Dataset-v2.web3-crypto-jobs-scraper-sample-data
Web3 & Crypto Jobs Scraper
Scrape crypto, blockchain and Web3 job listings from web3.career with full pagination — job title, company, tags and apply link. Schedule it to monitor crypto-industry hiring.
What the actor scrapes
🪙 Web3 & Crypto Jobs Scraper — Blockchain Job Listings to JSON/CSV/Excel Scrape crypto, blockchain and Web3 job listings from web3.career — one of the largest dedicated Web3 job boards. This web3.career scraper paginates the entire board… See the full description on the dataset page: https://huggingface.co/datasets/logiover/web3-crypto-jobs-scraper-sample-data.web3-qa-datasetcrypto-web3-ai-agents-2026
⚡ Crypto, Web3 & Autonomous Financial AI Agents Dataset (2023–2026)
This dataset contains 100 strictly domain-filtered research papers focusing on Decentralized AI, Autonomous Financial Agents, Smart Contract Verification, Zero-Knowledge Proofs (ZKP), DeFi, and Multi-Agent Consensus (2023-2026).
📊 Features:
384-dimensional PyTorch Embeddings for Vector Search & Semantic Clustering
Strict Domain Verification: Passed 2-stage filtering (100% relevant to Crypto/AI)… See the full description on the dataset page: https://huggingface.co/datasets/beatsprom/crypto-web3-ai-agents-2026.web3-llm-instructions
web3-llm-instructions
Description
web3-llm-instructions is an instruction-following dataset focused on Web3, blockchain, and cryptocurrency concepts.
The dataset is designed for fine-tuning large language models (LLMs) to understand and generate responses about Web3 topics such as DeFi, NFTs, DAOs, smart contracts, and blockchain infrastructure.
Dataset Structure
Each record contains the following fields:
instruction: the task or question
input: optional… See the full description on the dataset page: https://huggingface.co/datasets/YosepMulia/web3-llm-instructions.web3-project-labelsData sources include:
https://thegrid.id
role-play-bench
Role-play Benchmark
A comprehensive benchmark for evaluating Role-play Agents in Chinese and English scenarios.
Dataset Summary
Role-play Benchmark is designed to evaluate Role-play Agents' ability to deliver immersive role-play experiences through Situated Reenactment. Unlike traditional benchmarks with verifiable answers, Role-play is fundamentally non-verifiable, e.g., there's no single "correct" response when a tsundere character is asked "Do you like me?". Instead… See the full description on the dataset page: https://huggingface.co/datasets/web3w/role-play-bench.Humanoid-Web3-Multilingual-Command-Dataset-v3web3cryptoweb3-identity-and-wallet-securityweb3mmluWeb3-Audit-Knowledge-Base