Skip to content

RajKumaar123/Angika-LowResource-Translator

Repository files navigation

Angika Low Resource Translator

This project is a neural machine translation system designed to translate Angika language sentences into English. Built with an encoder-decoder architecture featuring LSTM layers in TensorFlow Keras, it strives to bridge the gap between a rich cultural heritage and modern technology.

Angika is an endangered language spoken in the Anga region of India—primarily in southeastern Bihar and the Santhal Parganas division of Jharkhand—as well as parts of the Purnia district in Bihar and a small community in Nepal’s Morang district. This beautiful language, spoken by vibrant communities, faces the threat of fading away as younger generations gravitate toward dominant languages.

By developing automated translation tools, this project hopes to breathe new life into Angika, preserving its unique voice and stories for future generations. It is more than just technology; it is a heartfelt effort to keep alive a language that carries the identity, culture, and history of millions. Through translation, we aspire to make Angika accessible to a wider audience, fostering pride among its speakers and ensuring that this precious linguistic heritage is never lost.


Project Overview

  • Language Pair: Angika (source) → English (target)
  • Model: Encoder-Decoder with LSTM layers using TensorFlow Keras
  • Goal: To develop a machine translation system that helps preserve the Angika language by enabling automated translation into English.

Why This Matters

Angika is a linguistic treasure facing the risk of extinction. By creating tools that support translation, we are helping keep the language alive, empowering its speakers, and connecting Angika culture to the wider world. This project represents a step toward safeguarding intangible cultural heritage through AI and machine learning.


Contributing

Contributions, suggestions, and improvements are welcome to further this important initiative.


Features

  • Encoder-Decoder model for sequence-to-sequence translation
  • Inference models for real-time translation of Angika sentences
  • Sample input/output examples included
  • Easily extendable for other low-resource language translation tasks

Installation

  1. Clone the repository:
    git clone https://github.com/RajKumaar123/Angika-LowResource-Translator.git
    cd Angika-LowResource-Translator
    
    

About

Angika-LowResource-Translator is a neural machine translation model designed to translate between Angika and English. Built for low-resource language support, it leverages fine-tuned transformers to preserve cultural context and improve accessibility for underrepresented communities.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors