Transformer Model Translator

Transformer Model Translator is a Python project designed to translate text between languages using transformer-based models. The project provides flexibility to integrate with popular transformer architectures (e.g., BERT, GPT, T5) for language translation tasks.

Features

• Transformer-based language translation
• Configurable language pairs and hyperparameters
• Easy dependency management with Poetry

Requirements

• Python 3.6+
• Poetry (as the dependency manager)

Installation

1. Clone the repository:

   git clone https://github.com/your-username/transformer-model-translater.git
   cd transformer-model-translater

2. Install dependencies using Poetry:

   poetry install

3. Activate the environment:

   poetry shell

4. Run the main to start training

Configuration

The project uses a configuration file, config.yaml, to define key parameters for the translation task. This includes:

• Source and target languages: Defines the language pair for translation (e.g., English to French).
• Batch size: The number of samples processed in each training/testing step.
• Learning rate: The step size for model optimization.
• Number of epochs: The number of times the model will iterate over the training dataset.