README.md

metadata

language:
  - hi
  - en
pretty_name: d
size_categories:
  - n<1K

Hindi-English Translation Dataset

📌 Overview

This dataset consists of parallel Hindi-English text pairs intended for training and fine-tuning machine translation models. The data has been collected from various sources to ensure diversity in sentence structures and vocabulary.

📂 Dataset Structure

• Source Language: Hindi (hi)
• Target Language: English (en)
• Format: CSV / JSON / TXT
• Columns:
  • source_text: The original Hindi sentence
  • target_text: The corresponding English translation

📊 Data Collection

The dataset has been curated from multiple publicly available resources, including open datasets, literary texts, and crowd-sourced translations. Low-quality translations have been filtered out to improve accuracy.

🔥 Use Cases

• Training machine translation models
• Fine-tuning pre-trained models like IndicTrans
• Cross-lingual NLP research
• Evaluating Hindi-English translation quality

🚀 How to Use

Load the dataset in Python:

from datasets import load_dataset

dataset = load_dataset("your-huggingface-username/your-dataset-name")
print(dataset)