STEM & Non-STEM Q&A Datasets for LLM Training
Collection
9 items • Updated
Dataset Preview
The dataset viewer should be available soon. Please retry later.
file_name string | relative_path string | extension string | folder string | file_size int64 | binary unknown |
|---|---|---|---|---|---|
African American Studies.json | JSONs\African American Studies.json | .json | JSONs | 12,537 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Art & Design.json | JSONs\Art & Design.json | .json | JSONs | 15,430 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Art History.json | JSONs\Art History.json | .json | JSONs | 12,529 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Biology.json | JSONs\Biology.json | .json | JSONs | 13,998 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Calculus AB.json | JSONs\Calculus AB.json | .json | JSONs | 14,049 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Calculus BC.json | JSONs\Calculus BC.json | .json | JSONs | 12,072 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Chemistry.json | JSONs\Chemistry.json | .json | JSONs | 13,961 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
Comparative Government and Politics.json | JSONs\Comparative Government and Politics.json | .json | JSONs | 15,410 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
English Language.json | JSONs\English Language.json | .json | JSONs | 15,625 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
English Literature.json | JSONs\English Literature.json | .json | JSONs | 16,611 | [
123,
13,
10,
32,
32,
32,
32,
34,
115,
101,
99,
116,
105,
111,
110,
115,
34,
58,
32,
91,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
123,
13,
10,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
32,
34,
115,
101,
99,
95,
110,
... |
End of preview.
Multilingual STEM & Non-STEM Question Answering Dataset (6.7M+ Q&A)
Overview
The Multilingual STEM & Non-STEM Question Answering Dataset is one of the largest educational Q&A resources designed for AI, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), educational assistants, intelligent tutoring systems, and academic research.
The dataset contains over 6.7 million educational question-answer pairs spanning STEM, Non-STEM, competitive examinations, language learning, professional education, and general knowledge domains.
Every sample includes a question, multiple-choice options, correct answer, and detailed explanation, making the dataset suitable for both supervised fine-tuning and reasoning-focused AI systems.
The complete collection is available in both PDF and structured JSON formats.
Dataset Highlights
| Property | Value |
|---|---|
| Total Questions | 6.7M+ |
| English Questions | 2.5M+ |
| Total Tokens | 1.8B+ |
| Average Words per Question | ~210 |
| Formats | PDF, JSON |
| Languages | 12 |
| Domains | STEM + Non-STEM |
| Question Type | Multiple Choice |
| Includes Explanations | ✅ |
| Mathematical Equations | LaTeX & MathML |
| Embedded Educational Images | ✅ |
Supported Languages
| Language | Questions |
|---|---|
| English | 4,073,358 |
| Hindi | 1,645,977 |
| Telugu | 158,044 |
| Bengali | 145,979 |
| Arabic | 139,496 |
| Marathi | 135,971 |
| Tamil | 102,569 |
| Gujarati | 69,580 |
| Kannada | 56,976 |
| Punjabi | 20,114 |
| Odia | 18,465 |
| Malayalam | 14,705 |
Subject Coverage
The dataset covers a broad range of educational disciplines including:
STEM
- Mathematics
- Physics
- Chemistry
- Biology
- Engineering
- Medical Sciences
- Computer Science
- Information Technology
- Environmental Science
- Agriculture
- Nursing
- Ayurveda
- General Science
- Quantitative Aptitude
Non-STEM
- History
- Geography
- Economics
- Commerce
- Business Studies
- Law
- Sociology
- Political Science
- General Knowledge
- Communication
- Teaching
- Library Science
- Management
- English Language
- Hindi Language
- Arabic Language
- Islamic Education
- Art & Design
- Home Science
- Social Science
- Miscellaneous Educational Topics
Largest Categories
| Category | Questions |
|---|---|
| Miscellaneous | 1,070,040+ |
| General Knowledge | 831,800+ |
| General English | 477,000+ |
| Reasoning | 725,000+ |
| Quantitative Aptitude | 507,000+ |
| Mathematics | 438,000+ |
| Engineering | 390,000+ |
| Medical | 161,000+ |
| Physics | 167,000+ |
| Biology | 149,000+ |
| Chemistry | 139,000+ |
Data Format
Each record contains structured educational information similar to:
{
"question": "...",
"options": [
"...",
"...",
"...",
"..."
],
"answer": "...",
"explanation": "...",
"language": "English",
"subject": "Physics"
}
Features
- 6.7M+ verified educational Q&A pairs
- Detailed explanations for every answer
- Multi-language support
- STEM and Non-STEM coverage
- Competitive examination content
- Higher education material
- School curriculum
- College curriculum
- Government exam preparation
- Mathematical equations in LaTeX
- Mathematical equations in MathML
- Embedded educational illustrations
- PDF source documents
- Structured JSON files
- Ideal for instruction tuning
- Ready for Retrieval-Augmented Generation (RAG)
Potential AI Applications
- Large Language Model (LLM) Training
- Supervised Fine-Tuning (SFT)
- Instruction Tuning
- Retrieval-Augmented Generation (RAG)
- Educational Chatbots
- AI Tutors
- Question Answering Systems
- Reasoning Models
- Benchmark Creation
- Knowledge Distillation
- Curriculum Learning
- Educational Search Engines
- Adaptive Learning Systems
- AI Exam Preparation Platforms
Industries
- Artificial Intelligence
- Education Technology (EdTech)
- E-Learning
- Universities
- Schools
- Government Education
- Competitive Exam Platforms
- Research Organizations
- Knowledge Management
- Digital Libraries
- Publishing
- Enterprise AI
File Structure
Dataset/
│
├── PDFs/
│ ├── Biology.pdf
│ ├── Mathematics.pdf
│ ├── Engineering.pdf
│ └── ...
│
├── JSONs/
│ ├── Biology.json
│ ├── Mathematics.json
│ ├── Engineering.json
│ └── ...
Dataset Statistics
- Questions: 6.7M+
- English Questions: 2.5M+
- Languages: 12
- Subjects: 45+
- Average Question Length: ~210 words
- Total Tokens: 1.8 Billion+
- Formats: PDF & JSON
- Supports Images: Yes
- Supports Equations: LaTeX & MathML
Advantages
- Extremely large multilingual educational corpus
- Rich reasoning-based explanations
- Suitable for both language and vision-language models
- Covers school, university, and professional education
- Balanced STEM and Non-STEM content
- Ready for scalable AI training pipelines
- Well-structured JSON annotations
- Human-readable PDF references
Limitations
- Dataset is intended primarily for educational and research applications.
- Some subjects contain significantly more samples than others.
- Coverage varies across supported languages.
License
This dataset is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.
Users are free to use, modify, distribute, and build upon the dataset with appropriate attribution.
Citation
@dataset{multilingual_qa_dataset,
title={Multilingual STEM & Non-STEM Question Answering Dataset},
year={2026},
license={CC BY 4.0}
}
Conclusion
The Multilingual STEM & Non-STEM Question Answering Dataset provides one of the most comprehensive educational corpora currently available for AI development. With 6.7 million+ questions, 1.8 billion+ tokens, extensive multilingual coverage, detailed explanations, and support for mathematical notation and educational imagery, it serves as a high-quality resource for training next-generation Large Language Models, educational assistants, reasoning systems, and intelligent tutoring applications.
- Downloads last month
- 11