mahahindi-425k / README.md
ayan-sh003's picture
Update README.md
463abfc verified
metadata
license: mit
language:
  - hi
size_categories:
  - 100K<n<1M
task_categories:
  - question-answering

Overview

This dataset is a collection of Hindi language data, comprising three main categories:

  1. Questions and Answers
  2. Mathematics Problems and Solutions
  3. Coding Problems and Solutions

The dataset aims to provide a rich resource for natural language processing (NLP) tasks, including question answering, mathematical problem-solving, and coding-related queries and responses in the Hindi language.

Dataset Structure

Questions and Answers

Each entry in this category consists of a question in Hindi and its corresponding answer. This section can be used to train and evaluate models for various NLP tasks, such as question answering, information retrieval, and language understanding.

Mathematics Problems and Solutions

This category includes mathematical problems stated in Hindi, along with their detailed solutions. The problems cover a range of mathematical topics and difficulty levels, making this section useful for educational purposes and for developing models that understand and solve math problems in Hindi.

Coding Problems and Solutions

Entries in this category present coding problems described in Hindi and their solutions. This can be particularly valuable for training models on programming-related tasks, as well as for educational tools aimed at Hindi-speaking students and professionals in the field of computer science.

This dataset is an aggregate of following datasets

  • ai4bharat/indic-instruct-data-v0.1 (lmsys)
  • FreedomIntelligence/evol-instruct-hindi
  • Cognitive-Lab/Aya_Hindi (adversarial_qa)
  • projectbaraat/hindi-Mathematical

Thanks to all the authors