Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 1087180155 bytes, limit is 300000000 bytes Make sure that 1. the Parquet files contain a page index to enable random access without loading entire row groups2. otherwise use smaller row-group sizes when serializing the Parquet files
Error code:   TooBigContentError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.


license: other task_categories:

  • audio-classification
  • automatic-speech-recognition language:
  • en
  • hi
  • bn
  • ta
  • te
  • ml
  • mr
  • or
  • as
  • pa tags:
  • ncert,
  • education,
  • india,
  • upsc,
  • humanities,
  • recorded-lectures,
  • government-exams,
  • science,
  • long-form-audio pretty_name: NCERT Lectures India size_categories:
  • 1K<n<10K

NCERT Lectures India

Dataset Description

A large-scale collection of recorded lectures covering NCERT curriculum across Science and Humanities streams, specifically designed for government exam preparation including UPSC, SSC, and State PSC examinations. Produced by Prepp, India's largest government exam preparation platform, operated by Collegedunia Web Private Limited.

Dataset Summary

  • Total duration: 5,100 hours of recorded lectures
  • Content type: Structured curriculum lectures covering NCERT Science and Humanities
  • Exam relevance: UPSC Civil Services, SSC, State PSC, Railways, Banking, and all major government competitive examinations
  • Curriculum: Full NCERT coverage — Classes 6 through 12, Science and Humanities streams
  • Chapters available: History Ch-1, Ch-2, Ch-3 (samples)
  • Languages: Hindi and English primary; regional language variants available
  • Format: Audio/Video with structured chapter-by-chapter delivery

Sample Data

Three sample lectures are available in this repository:

  • History Chapter 1 — The Early Societies
  • History Chapter 2 — Early Economies and Empires
  • History Chapter 3 — Political and Economic History

Key Features

  • NCERT-aligned: Follows official NCERT curriculum chapter structure exactly — high-value signal for Indian education AI models
  • Government exam optimised: Content structured specifically for competitive exam preparation — includes emphasis patterns, important facts, and exam-relevant framing that general educational content lacks
  • Long-form audio: 5,100 hours of continuous structured lecture audio — one of the largest Indic long-form educational audio datasets available
  • Subject breadth: Covers History, Geography, Polity, Economics, Science, and Environment across multiple NCERT grades
  • High information density: Unlike casual educational content, government exam lectures are dense with factual content — valuable for knowledge-intensive LLM training

Intended Uses

  • Training automatic speech recognition (ASR) models for Hindi and Indic languages in educational domain
  • Long-form audio understanding model development
  • Knowledge-intensive question answering model training
  • Indian history, geography, and polity domain model fine-tuning
  • Government exam preparation AI development
  • Curriculum-structured audio dataset for educational AI research

Why This Dataset Is Unique

Government exam preparation content is structurally different from general educational content. It is optimised for information retention, fact density, and conceptual clarity under exam conditions. No comparable large-scale structured dataset exists for Indian government exam preparation content in audio format.

Data Collection and Rights

All content is proprietary, produced by Prepp's in-house faculty team of government exam subject matter experts. Content is curriculum-mapped to NCERT chapter structure and ethically sourced under work-for-hire agreements. Full dataset licensing is available for commercial AI training purposes.

Licensing and Commercial Access

This repository contains sample data only. The full dataset of 5,100 hours of NCERT lecture recordings is available for commercial AI training licensing.

For licensing inquiries contact: Ankit Dubey — Head of AI Data Partnerships, Collegedunia ankit.dubey@collegedunia.com

Dataset Curator

Collegedunia Web Private Limited | Prepp

Gurugram, Haryana, India

Downloads last month
9