datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Code-Reasoning
Code-Reasoning
Dataset Description
Code problem-solving data with reasoning, direct-answer, and task-synthesis subsets. This repository is part of the K2 Horizon collection.
The repository is organized into multiple subsets. Every subset has a train split backed by Parquet shards, which supports Dataset Viewer inspection and streaming access.
K2 Horizon Dataset Series
Dataset repository
Focus
Subsets
IFM/TxT360-v2
Web and… See the full description on the dataset page: https://huggingface.co/datasets/IFM/Code-Reasoning.LiteCoder-Terminal-RL-preview
LiteCoder-Terminal-RL-preview
Paper | Code | Blog Post
This dataset contains 602 standardized Harbor terminal environments and was released as part of the paper LiteCoder-Terminal: Scaling Long-Horizon Terminal Environments for Learning Language Agents.
Unlike static text-only instructions, these environments are fully executable and are designed to support the training of terminal-based agents.
Environment Generation Pipeline
The lack of high-quality, executable… See the full description on the dataset page: https://huggingface.co/datasets/Lite-Coder/LiteCoder-Terminal-RL-preview.Ling-Coder-SFT
🤗 Hugging Face
🤖 ModelScope
🖥️ GitHub
Ling-Coder Dataset
The Ling-Coder Dataset comprises the following components:
Ling-Coder-SFT: A subset of SFT data used for training Ling-Coder Lite, containing more than 5 million samples.
Ling-Coder-DPO: A subset of DPO data used for training Ling-Coder Lite, containing 250k samples.
Ling-Coder-SyntheticQA: A subset of synthetic data used for annealing training of Ling-Coder Lite, containing more… See the full description on the dataset page: https://huggingface.co/datasets/inclusionAI/Ling-Coder-SFT.X-Coder-SFT-376k
X-Coder: Advancing Competitive Programming with Fully Synthetic Tasks, Solutions, and Tests
Dataset Overview
X-Coder-SFT-376k is a large-scale, fully synthetic dataset for advancing competitive programming.
The dataset comprises 4 subsets with a total of 887,321 synthetic records across 423,883 unique queries.
It is designed for supervised fine-tuning and suitbale for cold start to train code reasoning foundations.
X-Coder-SFT-376k is curated by sota reasoning models.… See the full description on the dataset page: https://huggingface.co/datasets/IIGroup/X-Coder-SFT-376k.github-codereview
Code Review Dataset
A large-scale dataset of the best human-written code reviews from top GitHub repositories.
Each row captures a moment where a human code reviewer left an inline comment on a pull request, and the author subsequently modified the code in response.
The dataset also includes negative examples — code from the same PRs that passed review without comments — to help models learn when code is acceptable.
This provides a natural signal for training models to:
Generate… See the full description on the dataset page: https://huggingface.co/datasets/ronantakizawa/github-codereview.Ling-Coder-SyntheticQA
🤗 Hugging Face
🤖 ModelScope
🖥️ GitHub
Ling-Coder Dataset
The Ling-Coder Dataset comprises the following components:
Ling-Coder-SFT: A subset of SFT data used for training Ling-Coder Lite, containing more than 5 million samples.
Ling-Coder-DPO: A subset of DPO data used for training Ling-Coder Lite, containing 250k samples.
Ling-Coder-SyntheticQA: A subset of synthetic data used for annealing training of Ling-Coder Lite, containing more… See the full description on the dataset page: https://huggingface.co/datasets/inclusionAI/Ling-Coder-SyntheticQA.Luau-Coder-1.0-Preview-SFT
Luau Coder 1.0 Preview SFT 🦭
This dataset is exceptionally high-quality supervised fine-tuning conversations for a highly capable coding model in Roblox Luau domain.
It prioritize technical correctness, useful engineering judgment, realistic interaction, and efficient explanations over output volume.
This dataset includes & covering:
Multi-turns (4-10 turns)
Dynamic CoT (length)
Dynamic Interleaved Reasoning
Long Context Session
Q/A
Review
Debugging
Bug Fix… See the full description on the dataset page: https://huggingface.co/datasets/khtsly/Luau-Coder-1.0-Preview-SFT.CodeRM-UnitTest
Dataset Description
CodeRM-UnitTest dataset originates from the paper: Dynamic Scaling of Unit Tests for Code Reward Modeling available on arXiv.
You can visit the homepage to learn more about the paper.
It is a curated collection of high-quality synthetic Python unit tests, derived from two prominent code instruction tuning
datasets: CodeFeedback-Filtered-Instruction and the training
set of TACO. This dataset is used for training
CodeRM-8B, a small yet powerful unit test… See the full description on the dataset page: https://huggingface.co/datasets/KAKA22/CodeRM-UnitTest.Qwen3-Coder-Next-Open-Code-SFT
Dataset Card — OpenCode Rejection Sampling
Overview
This dataset contains high-quality code reasoning data for training language models on competitive programming tasks. It is produced via rejection sampling with Qwen3-Coder-Next, which would generate multiple candidate solutions per problem, each candidate is executed against test cases in a sandboxed environment, and the results are used to build two complementary training datasets:
SFT dataset (49,374 examples)… See the full description on the dataset page: https://huggingface.co/datasets/zake7749/Qwen3-Coder-Next-Open-Code-SFT.Qwen3-Coder-Next-OpenCode-Preference
Dataset Card — OpenCode Rejection Sampling (Preference)
Overview
This dataset contains 10,920 preference pairs for preference-based training (DPO, KTO, SimPO, ORPO, etc.) on competitive programming tasks. Each pair consists of:
Chosen: a candidate solution that passes 100% of test cases
Rejected: a candidate solution that fails, with a fine-grained rejection type label
Pairs are produced via rejection sampling with Qwen3-Coder-Next: 8 candidate solutions are… See the full description on the dataset page: https://huggingface.co/datasets/zake7749/Qwen3-Coder-Next-OpenCode-Preference.WestGenesis-Coder-SFT-100M
Dataset Overview
WestGenesis-Coder-Dataset is a meticulously curated coding dataset designed specifically for instruction-based model tuning and fine-tuning of existing models with enhanced code generation capabilities. This represents one of the largest and most comprehensively filtered corpora of publicly available coding data on the Hugging Face platform, with a non-thinking approach that emphasizes direct, concise code outputs for rapid model training.
Key… See the full description on the dataset page: https://huggingface.co/datasets/isthatshan/WestGenesis-Coder-SFT-100M.Rust-Coder
Rust-Coder
Rust-Coder is a comprehensive text dataset designed for Rust programming language learning. It contains 12,000 unique samples focusing on distinct Rust concepts, code snippets, and explanations.
Dataset Structure
Each sample consists of:
id: A unique UUID.
instruction: A prompt or question about a Rust concept.
code: An idiomatic Rust code snippet.
explanation: A detailed explanation of the concept and code.
category: The high-level Rust category (e.g.… See the full description on the dataset page: https://huggingface.co/datasets/Convence/Rust-Coder.Ling-Coder-SFT-English-Clean
Ling-Coder-SFT-English-Clean
A cleaned, English-only version of inclusionAI/Ling-Coder-SFT — one of the largest open-source coding instruction datasets (~5.1M samples). Split by programming language for easy access.
Curated by MnemicAI
Origin Story
While building our Mnemic COCM-COT training pipeline — a multi-language coding instruction dataset with stratified topic sampling — we discovered that 11.44% of Ling-Coder-SFT contains Chinese/CJK characters mixed into what… See the full description on the dataset page: https://huggingface.co/datasets/MnemicAI/Ling-Coder-SFT-English-Clean.coder-pretrain-60gb
Coding LLM Pretraining Corpus
Built with build_pretrain_dataset.ipynb + finalize_local.py (DuckDB).
Composition
Source
Collected rows
Collected size
Final rows
Final size
code
8,415,197
30.00 GB
8,413,850
29.88 GB
web
3,492,492
15.00 GB
3,390,372
14.50 GB
math
659,720
3.00 GB
658,089
2.95 GB
wiki
519,584
2.00 GB
519,535
1.99 GB
docs
4,504,069
10.00 GB
4,289,545
9.74 GB
Total final corpus: 59.06 GB of raw text
(17,271,391 documents) across… See the full description on the dataset page: https://huggingface.co/datasets/Aniket200325/coder-pretrain-60gb.combined_coder_pythonCombining smaller python code datasets into a larger one.
Changed format to system, instruction, output.
Built from:
dataset1: nickrosh/Evol-Instruct-Code-80k-v1
dataset2: ehartford/dolphin-coder
dataset3: iamtarun/python_code_instructions_18k_alpaca
dataset4: iamtarun/python_code_instructions_18k_alpaca
dataset5: Vezora/Tested-22k-Python-Alpaca
dataset6: mlabonne/Evol-Instruct-Python-26k
dataset7: KrisPi/PythonTutor-Evol-1k-DPO-GPT4_vs_35
dataset8:… See the full description on the dataset page: https://huggingface.co/datasets/jtatman/combined_coder_python.Ling-Coder-SFT
🤗 Hugging Face
🤖 ModelScope
🖥️ GitHub
Ling-Coder Dataset
The Ling-Coder Dataset comprises the following components:
Ling-Coder-SFT: A subset of SFT data used for training Ling-Coder Lite, containing more than 5 million samples.
Ling-Coder-DPO: A subset of DPO data used for training Ling-Coder Lite, containing 250k samples.
Ling-Coder-SyntheticQA: A subset of synthetic data used for annealing training of Ling-Coder Lite, containing more… See the full description on the dataset page: https://huggingface.co/datasets/dinhuclab/Ling-Coder-SFT.OpenMathReasoning-mini-ko
OpenMathReasoning-mini Korean (한국어 번역)
이 데이터셋은 unsloth/OpenMathReasoning-mini 데이터셋을 한국어로 번역한 버전입니다.
원본 데이터셋은 nvidia/OpenMathReasoning 에서 파생되었습니다.
데이터셋 정보
항목
내용
원본 데이터셋
unsloth/OpenMathReasoning-mini
원본 출처
nvidia/OpenMathReasoning
라이선스
CC-BY-4.0
행 수
19,252
언어
한국어 (Korean), 영어 (English)
번역 대상 컬럼
expected_answer, problem, generated_solution
라이선스 및 귀속
이 데이터셋은 원본 데이터셋인 nvidia/OpenMathReasoning의 CC-BY-4.0 라이선스를 따릅니다.
원저작자: NVIDIA… See the full description on the dataset page: https://huggingface.co/datasets/neuralfoundry-coder/OpenMathReasoning-mini-ko.code-review-bench
Code Review Bench
A paired online-offline benchmark for AI code review.
Splits
online — Stratified sample of 1,135 bot-reviewed PRs, scraped from open-source Github repositories and scored by the online benchmark (15 tools, Feb–Apr 2026).
offline — 136 expert-curated golden issues across 50 PRs (5 repositories).
Provenance
The offline golden issues extend the 50-PR benchmark originally created by Greptile (2025) and refined by Augment (2025). Our… See the full description on the dataset page: https://huggingface.co/datasets/code-review-bench/code-review-bench.Qwen2.5-Coder-0.5B-Flutter-steps-eval
Qwen2.5-Coder-0.5B Flutter — Steps Mode — Validation Results
Dataset Summary
Held-out evaluation results for bbidpa/Qwen2.5-Coder-0.5B-Flutter-steps,
a fine-tune of Qwen2.5-Coder-0.5B for editing Flutter/Dart source files. In steps
mode, the model is given an existing file and an edit instruction and generates a
sequence of localized search/replace edit actions, each mechanically applied to the
current file state before the next action is generated, until the… See the full description on the dataset page: https://huggingface.co/datasets/bbidpa/Qwen2.5-Coder-0.5B-Flutter-steps-eval.Code-Regression
Code-Regression
Paper | GitHub Repository | Project Page
A unified regression dataset collated from three sources (APPS, KBSS, CDSS) along with our own custom profiling for training and evaluating regression models that map code strings to a target metric. This dataset supports "code-to-metric regression," which involves predicting numeric outcomes of code executions using Regression Language Models (RLM), as described in the linked paper.
Link for Graph-Regression dataset:… See the full description on the dataset page: https://huggingface.co/datasets/akhauriyash/Code-Regression.Qwen2.5-Coder-0.5B-Flutter-direct-eval
Qwen2.5-Coder-0.5B Flutter — Direct Mode — Validation Results
Dataset Summary
Held-out evaluation results for bbidpa/Qwen2.5-Coder-0.5B-Flutter-direct,
a fine-tune of Qwen2.5-Coder-0.5B for editing Flutter/Dart source files. In direct
mode, the model is given an existing file and an edit instruction and generates the
complete modified file in a single forward pass (as opposed to the steps /
iterative diff-based mode — see the sibling dataset… See the full description on the dataset page: https://huggingface.co/datasets/bbidpa/Qwen2.5-Coder-0.5B-Flutter-direct-eval.Rust-Coder
Rust-Coder
Rust-Coder is a comprehensive text dataset designed for Rust programming language learning. It contains 12,000 unique samples focusing on distinct Rust concepts, code snippets, and explanations.
Dataset Structure
Each sample consists of:
id: A unique UUID.
instruction: A prompt or question about a Rust concept.
code: An idiomatic Rust code snippet.
explanation: A detailed explanation of the concept and code.
category: The high-level Rust category (e.g.… See the full description on the dataset page: https://huggingface.co/datasets/gubernac/Rust-Coder.Persian-Wikipedia-Corpus
Overview
This dataset is derived from the Persian Wikipedia Corpus project, which contains parsed articles from the Persian Wikipedia. The original data has been converted into a more accessible format and made available through the HuggingFace datasets library.
Usage
from datasets import load_dataset
dataset = load_dataset("codersan/Persian-Wikipedia-Corpus")
Persian-Wikipedia-Corpus
A complete copy of Persian Wikimedia pages, The dataset contains articles… See the full description on the dataset page: https://huggingface.co/datasets/codersan/Persian-Wikipedia-Corpus.Qwen3-Coder-Next-Open-Code-SFT
Dataset Card — OpenCode Rejection Sampling
Overview
This dataset contains high-quality code reasoning data for training language models on competitive programming tasks. It is produced via rejection sampling with Qwen3-Coder-Next, which would generate multiple candidate solutions per problem, each candidate is executed against test cases in a sandboxed environment, and the results are used to build two complementary training datasets:
SFT dataset (49,374 examples)… See the full description on the dataset page: https://huggingface.co/datasets/derekib/Qwen3-Coder-Next-Open-Code-SFT.Ling-Coder-DPO
🤗 Hugging Face
🤖 ModelScope
🖥️ GitHub
Ling-Coder Dataset
The Ling-Coder Dataset comprises the following components:
Ling-Coder-SFT: A subset of SFT data used for training Ling-Coder Lite, containing more than 5 million samples.
Ling-Coder-DPO: A subset of DPO data used for training Ling-Coder Lite, containing 250k samples.
Ling-Coder-SyntheticQA: A subset of synthetic data used for annealing training of Ling-Coder Lite, containing more… See the full description on the dataset page: https://huggingface.co/datasets/inclusionAI/Ling-Coder-DPO.code-review
CODE_REVIEW
A preference dataset for CODE_REVIEW, harvested from real, human-labelled sources and curated by an automated harvesting harness with an LLM quality gate.
Format
Standard preference / DPO schema — each row:
column
meaning
prompt
the request (originally code)
chosen
the human-preferred response
rejected
a worse response to the same prompt
source
the dataset/URL the row was harvested from
Splits
80/10/10 train /… See the full description on the dataset page: https://huggingface.co/datasets/316usman/code-review.codereview-bench
CodeReview-Bench
A benchmark for evaluating models on two code review tasks, curated from ronantakizawa/github-codereview.
Tasks
1. Code Editing
Given code and a reviewer comment, apply the requested change.
Input: before_code, reviewer_comment, language, diff_context
Target: after_code
from datasets import load_dataset
ds = load_dataset("ronantakizawa/codereview-bench", "code-editing")
example = ds["test"][0]
prompt = f"""Apply the following review comment… See the full description on the dataset page: https://huggingface.co/datasets/ronantakizawa/codereview-bench.eval-trajs-1930-coder
⚠️ DO NOT TRAIN ON THIS — THIS IS TEST DATA ⚠️
These trajectories are evaluation outputs on SWE-bench-Verified.
Training on them (directly, or via distillation, rejection sampling,
or any form of preference data) contaminates the benchmark.
Every row carries a warning column repeating this notice.
Talkie 1930 / Web — SFT eval trajectories
Mini-SWE-Agent trajectories from the 2e-5 SFT runs of talkie-lm/talkie-1930-13b and talkie-lm/talkie-web-13b.
Graded by the SWE-bench… See the full description on the dataset page: https://huggingface.co/datasets/ricdomolm/eval-trajs-1930-coder.korean-legal-instruction-sample
Korean Legal Instruction Dataset (한국어 법률 지시학습 데이터셋)
데이터셋 개요
이 데이터셋은 대한민국 법률 도메인에 특화된 sLLM 지시학습(Instruction Tuning)용 데이터셋입니다.
AIHub에서 제공하는 16종의 법률 관련 데이터를 통합하여 현대 LLM 지시학습 포맷으로 가공하였습니다.
주요 특징
총 데이터 수: 약 233,000건
언어: 한국어
포맷: ChatML/Alpaca 호환 대화 형식
도메인: 법률 (민사, 형사, 행정, 지식재산권, 계약 등)
데이터 구조
각 데이터 샘플은 다음과 같은 구조를 가집니다:
{
"id": "고유 식별자",
"category": "카테고리명",
"source": "원본 데이터 출처",
"system": "시스템 프롬프트",
"instruction": "사용자 질문/지시",
"output": "AI… See the full description on the dataset page: https://huggingface.co/datasets/neuralfoundry-coder/korean-legal-instruction-sample.Eve-Secure-Coder
Eve Secure Coder Dataset
Dataset Description
Eve Secure Coder is a composite dataset curated to fine-tune Qwen3-4b-Heretic into a highly capable, security-conscious coding assistant with a distinct personality and no refusals.
The primary goal of this dataset is to bridge the gap between sterile, robotic coding assistants and engaging, conversational AI, without sacrificing technical accuracy or security. By combining high-quality coding archives, vulnerability detection… See the full description on the dataset page: https://huggingface.co/datasets/TitleOS/Eve-Secure-Coder.