karanverma19/multilingual-indian-instruction-dataset

---
license: apache-2.0
pretty_name: Multilingual Context-Aware Instruction Dataset for Indian Use Cases
---

# Multilingual Context-Aware Instruction Dataset for Indian Use Cases

## Overview
This dataset adapts multilingual instruction data to better support Indian languages such as Hindi, Hinglish, and Punjabi.

## Purpose
It focuses on real-world use cases including:
- Banking
- Government services
- Education
- Daily life interactions

## Key Features
- Multilingual (English, Hindi, Hinglish, Punjabi)
- Context-aware structure (instruction + context + response)
- Domain-specific tagging (banking, finance, government, etc.)
- Designed for real-world AI applications

## Motivation
Many existing datasets are generic and lack localized context. This dataset improves usability for low-resource and multilingual environments by introducing contextual understanding.

## Inspiration
This work is inspired by large multilingual datasets like Aya but focuses on practical, localized AI applications.

## Data Source & Adaptation
This dataset is inspired by multilingual instruction datasets such as Aya and adapts the structure to focus on Indian languages and real-world use cases.