📖 Machine Translation (English to Arabic)

This project demonstrates how to use Hugging Face libraries to apply the Helsinki-NLP opus-mt-en-ar model for English-to-Arabic translation. It includes:

• Loading the CoVoST2 EN-AR dataset
• Setting up the model and tokenizer
• Translating text using a pre-trained model
• Evaluating results with BLEU scores

---

Features

Model and Tokenizer Setup

• Load Pre-trained Model: Load the Helsinki-NLP opus-mt-en-ar model.
• Initialize Tokenizer: Set up the tokenizer for text processing.

Text Translation

• Tokenize Input Text: Prepare the input text for translation.
• Generate Translations: Use the model to translate the text.
• Decode Output: Convert the model's output to readable text.

Evaluation

• Compute BLEU Scores: Compare model translations with reference translations to measure quality.

Demonstration

• Example Translations: Showcase the model's capabilities with example sentences.
• Advantages: Highlight the benefits of using a pre-trained model.

---

Getting Started

Prerequisites

• Python 3.x
• Hugging Face Transformers library
• Datasets library

Installation

1. Clone the Repository:

   git clone https://github.com/ManwanMaro999/Machine-Translation.git
   cd Machine-Translation

2. Install Dependencies:

   pip install -r requirements.txt

Project Video

Machine.Translation.project.mp4