datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
hermes-function-calling-v1
Hermes Function-Calling V1
This dataset is the compilation of structured output and function calling data used in the Hermes 2 Pro series of models.
This repository contains a structured output dataset with function-calling conversations, json-mode, agentic json-mode and structured extraction samples, designed to train LLM models in performing function calls and returning structured output based on natural language instructions. The dataset features various conversational scenarios… See the full description on the dataset page: https://huggingface.co/datasets/NousResearch/hermes-function-calling-v1.Magicoder-Evol-Instruct-110KA decontaminated version of evol-codealpaca-v1. Decontamination is done in the same way as StarCoder (bigcode decontamination process).
dolmino-mix-1124
DOLMino dataset mix for OLMo2 stage 2 annealing training.
Mixture of high-quality data used for the second stage of OLMo2 training.
Source Sizes
Name
Category
Tokens
Bytes (uncompressed)
Documents
License
DCLM
HQ Web Pages
752B
4.56TB
606M
CC-BY-4.0
Flan
HQ Web Pages
17.0B
98.2GB
57.3M
ODC-BY
Pes2o
STEM Papers
58.6B
413GB
38.8M
ODC-BY
Wiki
Encyclopedic
3.7B
16.2GB
6.17M
ODC-BY
StackExchange
CodeText
1.26B
7.72GB
2.48M
CC-BY-SA-{2.5, 3.0, 4.0}… See the full description on the dataset page: https://huggingface.co/datasets/allenai/dolmino-mix-1124.medical-o1-reasoning-SFT
News
[2025/04/22] We split the data and kept only the medical SFT dataset (medical_o1_sft.json). The file medical_o1_sft_mix.json contains a mix of medical and general instruction data.
[2025/02/22] We released the distilled dataset from Deepseek-R1 based on medical verifiable problems. You can use it to initialize your models with the reasoning chain from Deepseek-R1.
[2024/12/25] We open-sourced the medical reasoning dataset for SFT, built on medical verifiable problems and an… See the full description on the dataset page: https://huggingface.co/datasets/FreedomIntelligence/medical-o1-reasoning-SFT.security-auditsA collection of agent traces generated with Swival (not Claude Code, despite what the HF interface currently shows), an agent designed for open-source models.
These traces focus on security audits of opensource software.
Sharing traces with Swival
Swival can export full conversation traces with --trace-dir, which writes one <session_id>.jsonl file per session:
swival "Fix the login bug" --trace-dir traces/
Those JSONL files use Swival's Claude Code compatible trace export, and… See the full description on the dataset page: https://huggingface.co/datasets/jedisct1/security-audits.SkyPile-150B
SkyPile-150B
Dataset Summary
SkyPile-150B is a comprehensive, large-scale Chinese dataset specifically designed for the pre-training of large language models. It is derived from a broad array of publicly accessible Chinese Internet web pages. Rigorous filtering, extensive deduplication, and thorough sensitive data filtering have been employed to ensure its quality. Furthermore, we have utilized advanced tools such as fastText and BERT to filter out low-quality data.
The… See the full description on the dataset page: https://huggingface.co/datasets/Skywork/SkyPile-150B.OpenMathInstruct-1
OpenMathInstruct-1
OpenMathInstruct-1 is a math instruction tuning dataset with 1.8M problem-solution pairs
generated using permissively licensed Mixtral-8x7B model.
The problems are from GSM8K
and MATH training subsets and the solutions
are synthetically generated by allowing Mixtral model to use a mix of text reasoning and
code blocks executed by Python interpreter.
The dataset is split into train and validation subsets that we used in the ablations experiments.
These two subsets… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/OpenMathInstruct-1.evol-codealpaca-v1
Evolved codealpaca
Updates:
2023/08/26 - Filtered results now only contain pure english instruction and removed any mentioned of trained by OAI response
Median sequence length : 471
We employed a methodology similar to that of WizardCoder, with the exception that ours is open-source. We used the gpt-4-0314 and gpt-4-0613 models to augment and answer each response, with the bulk of generation handled by gpt-4-0314.
The aim of this dataset is twofold: firstly, to facilitate the… See the full description on the dataset page: https://huggingface.co/datasets/theblackcat102/evol-codealpaca-v1.SWE-Fixer-Train-110K
SWE-Fixer: Training Open-Source LLMs for Effective and Efficient GitHub Issue Resolution
📃 Paper |
🚀 GitHub
SWE-Fixer is a simple yet effective solution for addressing real-world GitHub issues by training open-source LLMs. It features a streamlined retrieve-then-edit pipeline with two core components: a code file retriever and a code editor.
This repo holds the data SWE-Fixer-Train-110K we curated for SWE-Fixer training.
For more information, please visit our project page.… See the full description on the dataset page: https://huggingface.co/datasets/internlm/SWE-Fixer-Train-110K.dolma3_mix-150B-1025
Dolma 3 Sample: 150B Mix
Dataset Sources
Sample of data for 1Bx5C and 7Bx1B. For the full Dolma 3 pool, see: https://huggingface.co/datasets/allenai/dolma3
Source
Type
Tokens
Documents
Common Crawl
Web pages
121B (76.9%)
84.5M
olmOCR Science PDFs
Academic documents
19.9B (12.6%)
2.25M
Stack-Edu (Rebalanced)
GitHub code
11.1B (7.06%)
14.3M
arXiv
Papers with LaTeX
1.29B (0.82%)
247K
FineMath 3+
Math web pages
4.10B (2.60%)
2.57M
Wikipedia & Wikibooks… See the full description on the dataset page: https://huggingface.co/datasets/allenai/dolma3_mix-150B-1025.soc-builder-rtl-v1
SoC Builder RTL Dataset — v1 (Experiment Release)
A reproducible, machine-generated corpus of synthesizable System-on-Chip (SoC) RTL designs for machine learning on hardware: RTL representation learning today, and — as the corpus grows — netlist, timing, and placement prediction. Every design is a complete, hierarchical, lint-clean Verilog SoC assembled from real open-source IP — RISC-V CPU cores, network-on-chip (NoC) interconnects, accelerators, peripherals, memories and… See the full description on the dataset page: https://huggingface.co/datasets/hasankursun/soc-builder-rtl-v1.novel-agent-sft-dataset
All Novel Can Be Galgame — 完整数据集
中文小说叙事理解项目的完整数据集。包含 669 本中文小说的原始文本、标注和训练数据，用于训练叙事 Agent 系统。
项目地址：https://github.com/lin1753/novel2galgame
训练代码仓库：https://github.com/lin1753/novel-agent
数据规模
目录
文件数
大小
说明
training/
52
689 MB
训练用 SFT 数据 (JSONL)
raw-books/
671
327 MB
669 本原始小说
processed/
39,842
1.2 GB
按章节预处理文本
annotations/
1,626
1 MB
原始标注文件
合计
42,191
2.2 GB
目录结构
datasets/
├── training/
│ ├── base-sft/… See the full description on the dataset page: https://huggingface.co/datasets/mikuhhn1239/novel-agent-sft-dataset.pii-masking-openpii-1.5m
OpenPII 1.5M: Multilingual PII Masking Dataset (Asia Pacific Extension)
📖 More information: www.ai4privacy.com/datasets/pii-masking-3m-asia-pacific
Overview
The OpenPII 1.5M dataset extends OpenPII 1M
with a new Asia Pacific corpus, bringing global coverage to 30 languages
across Europe, Americas, and Asia Pacific.
This is the flagship release of the PII-Masking-3M family, the world's
largest open multilingual PII masking corpus. Built to advance open… See the full description on the dataset page: https://huggingface.co/datasets/ai4privacy/pii-masking-openpii-1.5m.Fable-5.1-Max-Reasoning-Filtered-10000x
Dataset Description
This dataset contains 10,000 agentic coding and reasoning multi-turn high-quality traces generated by the new Fable 5.1 model using max reasoning effort.
It holds almost 500,000,000 tokens of step-by-step chain-of-thought programming across multiple complex domains.
It has also been deduplicated and heavily filtered to remove low-quality traces, keeping only high-quality traces.
Dataset Statistics
Metric
Value
Total Examples
10,000… See the full description on the dataset page: https://huggingface.co/datasets/MoreThought/Fable-5.1-Max-Reasoning-Filtered-10000x.counselbench-100
CounselBench-100
CounselBench-100 v3.2.5 is a synthetic legal-work benchmark with 100
authored matters across ten practice workflows. Every task has a natural employee
request, a 97-asset evidence room, twelve portfolio decisions, 5–9 supported
actions, 3–7 evidence holds, and a distinct deep multi-provider MCP trajectory.
The answer is not preclassified in the evidence. Each portfolio item requires an
immutable identity join, an operative-authority and revision lookup, a… See the full description on the dataset page: https://huggingface.co/datasets/SamuelChien821/counselbench-100.mimo-claude-code-traces-1k
MIMO Claude Code Traces
MIMO Claude Code Traces is a collection of coding-agent trajectories in a Claude Code-style environment. Each record contains a user coding task, the full multi-turn message trace, available tool schemas, assistant reasoning fields, tool calls, tool outputs, and metadata such as model name, category, duration, cost, token usage, and whether the trace used tools.
The traces were generated with mimo-v2.5-pro, MiMo's most capable model at the time of… See the full description on the dataset page: https://huggingface.co/datasets/choucsan/mimo-claude-code-traces-1k.kimi-cyber-reasoning
Kimi Cyber Reasoning
997 chain-of-thought records covering 13 cybersecurity disciplines and 4 systems engineering domains, distilled from the Kimi K3 reasoning model via API. Every record provides an explicit step-by-step <think> reasoning trace followed by a technical resolution, unified code diff fix, or structured tool invocation.
The dataset was curated as an anchor set for training, healing, and specializing compact reasoning models on systems security and tool calling… See the full description on the dataset page: https://huggingface.co/datasets/echel0nn1881/kimi-cyber-reasoning.Creative-Writing-High-Quality-1300x
Creative Writing - Part One (Shadow & Skeleton)
This dataset is designed to train Large Language Models (LLMs) in grounded creative writing by enforcing a "Think-Before-You-Write" methodology.
Methodology: Shadow & Skeleton
Most creative writing datasets train models to produce "vibes" or "cinematic descriptions" that often lack physical coherence or psychological depth. This dataset takes a different approach:
Shadow Prompts: We generated 1,000+ isomorphic… See the full description on the dataset page: https://huggingface.co/datasets/Crownelius/Creative-Writing-High-Quality-1300x.Light-R1-SFTData
Light-R1: Surpassing R1-Distill from Scratch* with $1000 through Curriculum SFT & DPO
*from models without long COT
technical report
GitHub page
Here are the two-stage SFT data we used to train Light-R1-32B.
Simply refer to stage1-76k.json and stage2-3k.json
Model
Trained From
Release Date
AIME24
AIME25
DeepSeek-R1-Distill-Llama-70B
Llama-3.3-70B-Instruct
25.1.20
70.0
54.1
DeepSeek-R1-Distill-Qwen-32B
Qwen2.5-32B
25.1.20
72.6
54.9
LIMO (32B)
Qwen2.5-32B-Instruct
25.2.4… See the full description on the dataset page: https://huggingface.co/datasets/qihoo360/Light-R1-SFTData.Nemotron-SFT-ARC-AGI-v1
Dataset Description:
Nemotron-SFT-ARC-AGI-v1 is a supervised fine-tuning (SFT) dataset of multi-turn agentic reasoning traces produced by open-weight large language models attempting to solve ARC-AGI visual-reasoning puzzles. Each ARC puzzle (a set of (input grid, output grid) demonstration pairs plus one or more test inputs, where grids are 2D integer arrays representing colors) is formatted as a text prompt and given to an agent powered by one of nine open-weight reasoning… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-SFT-ARC-AGI-v1.salesbench-100
SalesBench-100
SalesBench-100 is a synthetic long-horizon sales-agent benchmark with 100 original workflows across Salesforce, HubSpot, Gong, and a seeded evidence room. Each task begins with a high-level employee request and has its own authored causal rule and provider transition. Identity, operating facts, authority, governed policy, live-system indexes, and exceptions are separated so no mounted business asset publishes a selected option or precomputed change. Every task has… See the full description on the dataset page: https://huggingface.co/datasets/SamuelChien821/salesbench-100.pre_1929_books_filtered
Pre-1929 Books
Description
Books published in the US before 1929 passed into the public domain on January 1, 2024.
We used the bibliographic catalog Hathifiles produced by HathiTrust to identify digitized books which were published in the US before 1929.
The collection contains over 130,000 books digitized and processed by the Internet Archive on behalf of HathiTrust member libraries.
The OCR plain text files were downloaded directly from the Internet Archive website.… See the full description on the dataset page: https://huggingface.co/datasets/common-pile/pre_1929_books_filtered.Nemotron-RL-Agentic-Terminal-Pivot-v1
Dataset Description
The Nemotron-RL-Agentic-Terminal-Pivot-v1 dataset provides training samples for reinforcement learning of command-line ("terminal use") LLM agents with the terminus_judge environment in NeMo Gym.
Each record is a single agent decision point extracted from a successful agent trajectory on a terminal task:
responses_create_params.input — the prompt: the task instruction plus the terminal interaction history (prior agent actions and terminal outputs) up to the… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-RL-Agentic-Terminal-Pivot-v1.PKU-SafeRLHF-10K
Paper
You can find more information in our paper.
Dataset Paper: https://arxiv.org/abs/2307.04657
pii-masking-openpii-1m
OpenPII 1M — Multilingual PII Masking Dataset
Overview
The OpenPII 1M dataset is a large-scale, multilingual collection of 1,428,143 synthetic text examples with fine-grained PII (Personally Identifiable Information) annotations, spanning 23 European languages and 19 entity types.
Built to advance open research in privacy-preserving NLP, this dataset enables the development and benchmarking of Named Entity Recognition (NER) models, token classification pipelines… See the full description on the dataset page: https://huggingface.co/datasets/ai4privacy/pii-masking-openpii-1m.SCP-116K
Multimodal preview: SCP-VL. This extension of the SCP dataset family contains 41,828 English and Chinese scientific problem-solution pairs, each with one or more attached images. The current v0.1-preview release may contain missing information, incomplete or mismatched images, and solution errors. See the SCP-VL dataset card for details, known limitations, and loading instructions.
New Version Available: SCP-378K
A new version of this dataset has been released:
👉… See the full description on the dataset page: https://huggingface.co/datasets/EricLu/SCP-116K.Nemotron-SpecializedDomains-Finance-v1
Dataset Description
Nemotron-SpecializedDomains-Finance is a large-scale synthetic financial question-answering dataset designed to improve LLM performance on specialized financial reasoning and document comprehension tasks. The dataset comprises 326K+ high-quality Q&A pairs generated from SEC filings of S&P 500 companies spanning 2019-2024.
This dataset is ready for commercial use.
Overview
The dataset leverages template-based Synthetic Data Generation (SDG) to… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-SpecializedDomains-Finance-v1.alive-medical-imaging
ALIVE Medical Imaging QA Dataset
Lecture-derived question-answer corpus, retrieval index, and source
materials for the ALIVE (Avatar-Lecture Interactive Video Engine)
system. The dataset was built from 23 recorded lectures of an
undergraduate medical imaging course and is the corpus used to
fine-tune the ALIVE language model and to evaluate its retrieval and
answer-generation behavior.
Layout
huggingface/
├── data/ question-answer pairs (Alpaca-style… See the full description on the dataset page: https://huggingface.co/datasets/zabir1996/alive-medical-imaging.msi-corpus
Main Street Independent Corpus
The complete Main Street Independent archive as open data: 917 news articles and 0 opinion columns, each with full text and metadata. AI-generated news and opinion, dedicated to the public domain under CC0 — no rights reserved. Refreshed daily.
Load it
from datasets import load_dataset
ds = load_dataset("golfplan18/msi-corpus") # 'news' and 'opinion' splits
print(ds["news"][0]["headline"], ds["news"][0]["text"][:200])… See the full description on the dataset page: https://huggingface.co/datasets/golfplan18/msi-corpus.Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1
Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1
Dataset Description:
Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1 is an RL dataset for training and evaluating a tool-using agent's ability to resist Indirect Prompt Injection (IPI) attacks hidden inside tool-returned environment data. In each record, the agent receives a benign user request that requires calling a read tool whose output contains an adversarial instruction disguised as legitimate domain content… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1.