datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
FlashRAG_datasets
⚡FlashRAG: A Python Toolkit for Efficient RAG Research
FlashRAG is a Python toolkit for the reproduction and development of Retrieval Augmented Generation (RAG) research. Our toolkit includes 36 pre-processed benchmark RAG datasets and 16 state-of-the-art RAG algorithms.
With FlashRAG and provided resources, you can effortlessly reproduce existing SOTA works in the RAG domain or implement your custom RAG processes and components.
For more information, please view our GitHub repo… See the full description on the dataset page: https://huggingface.co/datasets/RUC-NLPIR/FlashRAG_datasets.medical_meadow_medical_flashcards
Dataset Card for Medical Flashcards
Dataset Summary
Medicine as a whole encompasses a wide range of subjects that medical students and graduates must master
in order to practice effectively. This includes a deep understanding of basic medical sciences, clinical knowledge,
and clinical skills. The Anki Medical Curriculum flashcards are created and updated by medical students and cover the
entirety of this curriculum, addressing subjects such as anatomy, physiology… See the full description on the dataset page: https://huggingface.co/datasets/medalpaca/medical_meadow_medical_flashcards.opengloss-v1.3-query-examples-flat
See also OpenGloss v2.1 (2026-09-07): a deeper release of 109,633 of these headwords — sense-level ids, four reading levels, sense-tagged examples with spans, a judged relation graph, and retrieval supervision — published as a 16-dataset family. v1.3 remains the broader headword list.
OpenGloss Query Examples v1.3 (Flattened)
Dataset Summary
OpenGloss Query Examples is a synthetic dataset of search queries generated for vocabulary
terms. Each term has multiple… See the full description on the dataset page: https://huggingface.co/datasets/mjbommar/opengloss-v1.3-query-examples-flat.HMMT_2025
Dataset Summary
This dataset comprises the questions, answers, and solutions from HMMT February 2025, all of which were extracted by OCR, converted to LaTeX, and manually verified by FlagEval Team.
Data Fields
Below one can find the description of each field in the dataset.
id (str): Index of the problem in the competition
problem (str): Full problem statement
answer (str): Ground-truth answer to the question
solution(str): Ground-truth solution to the question… See the full description on the dataset page: https://huggingface.co/datasets/FlagEval/HMMT_2025.flan2021-full
Task Name
FLAN-2021 -> 70
{
"ag_news_subset": 108497,
"ai2_arc/ARC-Challenge": 829,
"ai2_arc/ARC-Easy": 1927,
"aeslc": 13187,
"anli/r1": 15361,
"anli/r2": 41133,
"anli/r3": 91048,
"bool_q": 8343,
"cnn_dailymail": 259607,
"coqa": 6456,
"cosmos_qa": 22996,
"definite_pronoun_resolution": 1079,
"drop": 70045,
"fix_punct": 25690,
"gem/common_gen": 60936,
"gem/dart": 56724,
"gem/e2e_nlg": 30337,
"gem/web_nlg_en": 31899… See the full description on the dataset page: https://huggingface.co/datasets/aslawliet/flan2021-full.medical-meadow-medical-flashcards
Dataset Card for medical-meadow-medical-flashcards
This dataset originates from the medAlpaca repository.
The medical-meadow-medical-flashcards dataset is specifically used for models training of medical question-answering.
Dataset Details
Dataset Description
Each sample is comprised of three columns: instruction, input and output.
Language(s): English
Dataset Sources
The code from the original repository was adopted to post it here.
Repository:… See the full description on the dataset page: https://huggingface.co/datasets/flwrlabs/medical-meadow-medical-flashcards.KrynexAI-Dataset-Flash-Instruction
🧠 KrynexAI Dataset
English | Русский
📌 Overview
KrynexAI Dataset is a high-quality, synthetically expanded collection of 10,000+ instruction-response pairs designed for fine-tuning Large Language Models (LLMs).
The dataset covers a wide range of topics including:
💻 Programming (Python, algorithms, data structures)
🤖 AI & Machine Learning (neural networks, transformers, LLMs)
🔭 Science (physics, cosmology, biology, neuroscience)
🧠 Philosophy & Psychology… See the full description on the dataset page: https://huggingface.co/datasets/KrynexLabs/KrynexAI-Dataset-Flash-Instruction.k3-sft-cc0-flan
Dataset Card for K3 SFT CC0 FLAN
844-row Kimi K3 synthetic instruction-tuning shard built from DPI-traced CC0/public-domain
FLAN prompts in the Tülu mix. Four overlapping Hub configs expose different cohort
views; adaptive is the recommended default for quality-conscious SFT mixing.
Dataset Details
Curated by: Training Datasmith
Teacher: kimi-k3 via deltafin (local inference)
Languages: English prompts; translation pairs include German, Spanish, Czech, Igbo… See the full description on the dataset page: https://huggingface.co/datasets/Training-Datasmith/k3-sft-cc0-flan.qa-dataset-k1000
QA Dataset K1000 — The First Drop of Ink
Question-answering data with gold documents and distractor pools for long-context evaluation, accompanying The First Drop of Ink: Nonlinear Impact of Distracting Information in Long-Context Reasoning by Muhan Gao, Zih-Ching Chen, and Kuan-Hao Huang (ICML 2026).
Paper · Full text (v2) · Hugging Face paper page
The paper studies how the proportion of hard distractors affects performance at fixed context length. It reports a nonlinear… See the full description on the dataset page: https://huggingface.co/datasets/lab-flair/qa-dataset-k1000.glm-5.3-flash-function-calling
GLM-5.3-Flash Function Calling (synthetic)
A synthetic function-calling dataset generated with zai-org/GLM-5.3-Flash via Hugging Face Inference Providers.
513 examples in 8 domains: weather, calendar, finance, travel, e-commerce, devops, smart home, communication.
Categories: single-turn tool calls, parallel/multiple calls in one turn, multi-turn trajectories with tool results, and no-tool-needed turns.
Format: OpenAI-style — each row has tools (JSON-schema function… See the full description on the dataset page: https://huggingface.co/datasets/Rallex3/glm-5.3-flash-function-calling.math_mergedTraining dataset contains aime (excluding 2024), math/train, math/test, openai_math_splits/train, and KbsdJames/Omni-MATH/test. Total of 17521 lines of unique problems.
Testing dataset contains aime_24 and math500 (i.e. openai_math_splits/test). Total of 530 lines of unique problems.
math_merged_cot_solA dataset consists problems from flatlander1024/math_merged and cot solutions generated by Llama-3.1-8b-Instruct. The is_correct label indicates whether the solution is correct or not.
Number of lines: 13864, Overall correct rate: 57.3%
NuminaMath-CoT-filteredDataset that contains problems that appears in both QwQ-LongCoT-130K-cleaned and NuminaMath-CoT. There are approximately 100k problems where the solution is in plain-CoT manner.
flatbot-mini-35M-dataset
FlatBuild Demo Chat 10K Dataset
The FlatBuild Demo Chat 10K Dataset is the official conversational training dataset for FlatBuild and is used to train Flatbot-Mini-35M, the flagship demonstration language model of the Flatseek ecosystem.
The dataset showcases the complete workflow of building a conversational language model entirely from scratch, including:
dataset preparation
tokenizer training
chat data preprocessing
Transformer training
checkpoint export
GGUF conversion… See the full description on the dataset page: https://huggingface.co/datasets/flatseek/flatbot-mini-35M-dataset.flattened-MMLU
MMLU prompts (flattened)
This dataset is a flattened reformatting of the original MMLU benchmark, in the format expected by MASEval's MMLU benchmark with support for accelerated DISCO evaluation.
What this dataset is
mmlu_prompts_examples.json – JSON list of items with query, full_prompt, choices, gold, and optional example.
A flattened structure suitable for anchor-point evaluation and DISCO prediction pipelines.
What "flattened" means
The original MMLU has… See the full description on the dataset page: https://huggingface.co/datasets/arubique/flattened-MMLU.math_merged_cot_sol_hardA dataset consists the hard problems (aime + problems with level >= 5) from flatlander1024/math_merged and cot solutions generated by Qwen2.5-32B-Instruct. The is_correct label indicates whether the solution is correct or not.
Number of lines: 6940, Overall correct rate: 45.4%
Educational-Flashcards-for-Global-Learners
1. Educational-Flashcards-for-Global-Learners/README.md
Educational Flashcards Dataset
Overview
A comprehensive collection of 100 educational flashcards covering STEM, humanities, law, arts, and cultural topics. Curated with 70% Indian content, 25% European, and 5% other Asian perspectives to promote diverse knowledge representation.
Dataset Structure
{
"input": "Text description",
"output": {
"type": "flashcards",
"topic": "Subject name"… See the full description on the dataset page: https://huggingface.co/datasets/Srinivasmec26/Educational-Flashcards-for-Global-Learners.CLCC_v1CLCC is a Chinese Linguistics & Cognition Challenge dataset, which is created by FlagEval team (https://flageval.baai.ac.cn/#/home).
The details can be found in (https://flageval.baai.ac.cn/#/taskIntro?t=zh_oqa)
Evaluate
The results can evaluated by human or our judgeLLM (https://huggingface.co/FlagEval/flageval_judgemodel).
medical_meadow_medical_flashcards
Dataset Card for Medical Flashcards
Dataset Summary
Medicine as a whole encompasses a wide range of subjects that medical students and graduates must master
in order to practice effectively. This includes a deep understanding of basic medical sciences, clinical knowledge,
and clinical skills. The Anki Medical Curriculum flashcards are created and updated by medical students and cover the
entirety of this curriculum, addressing subjects such as anatomy, physiology… See the full description on the dataset page: https://huggingface.co/datasets/1jia/medical_meadow_medical_flashcards.opengloss-v1.2-query-examples-flat
OpenGloss Query Examples v1.2 (Flattened)
Dataset Summary
OpenGloss Query Examples is a synthetic dataset of search queries generated for vocabulary
terms. Each term has multiple query profiles covering different search intents and user personas,
making it ideal for training query generation, intent classification, and RAG systems.
This dataset contains flattened profile records (one per query).
It is derived from the OpenGloss
encyclopedic dictionary.
Key… See the full description on the dataset page: https://huggingface.co/datasets/mjbommar/opengloss-v1.2-query-examples-flat.cleand_flatlander1024_or_instruct_dedup元データ: https://huggingface.co/datasets/flatlander1024/or_instruct_dedup
使用したコード: https://github.com/LLMTeamAkiyama/0-data_prepare/tree/master/src/flatlander1024-or_instruct_dedup
データ件数: 2,600
平均トークン数: 1,340
最大トークン数: 3,086
合計トークン数: 3,484,377
ファイル形式: JSONL
ファイル分割数: 1
合計ファイルサイズ: 13.0 MB
加工内容：
データセットの初期設定と読み込み:
flatlander1024/or_instruct_dedup データセットを読み込み、Pandas DataFrameに変換しました。
answer 列のデータ型を文字列 (str) に変換しました。
NLTKのpunktとstopwordsデータをダウンロードしました（必要な場合）。
IDの付与:… See the full description on the dataset page: https://huggingface.co/datasets/LLMTeamAkiyama/cleand_flatlander1024_or_instruct_dedup.flatbot-micro-4M-dataset
FlatBuild Demo Chat 2.5K Dataset
The FlatBuild Demo Chat 2.5K Dataset is the official demonstration dataset for FlatBuild and is used to train Flatbot-Micro-4M, the introductory language model of the Flatseek ecosystem.
The dataset demonstrates the complete workflow of training a conversational language model entirely from scratch, including:
dataset preparation
tokenizer training
chat data preprocessing
Transformer training
checkpoint export
GGUF conversion
efficient inference… See the full description on the dataset page: https://huggingface.co/datasets/flatseek/flatbot-micro-4M-dataset.Safety_Alignment_BenchmarkGarud_puran_FlanT5_with_context
Garuda Purana Q&A for FLAN-T5
This dataset contains question-answer pairs from the Garuda Purana, with a summarization context for each pair generated by FLAN-T5.
Fields:
question: The input question in natural language.
answer: The answer to the question.
context: A short summary (generated by FLAN-T5) of the Q&A pair, usable as context or for semantic retrieval.
Intended Use:
Supervised fine-tuning for Question Answering, Retrieval, and Instruction-based LLMs.
The question… See the full description on the dataset page: https://huggingface.co/datasets/Binoddai/Garud_puran_FlanT5_with_context.