williampepple1/ibani-byt5-model

# Ibani Translator 🌍 A neural machine translation system for English to Ibani language using Hugging Face transformers. ## Features - **Neural Translation**: Fine-tuned ByT5 model for English to Ibani translation - **REST API**: FastAPI service with interactive documentation - **Training Pipeline**: Train custom models on your own Ibani data - **Batch Translation**: Translate multiple texts efficiently - **Model Hosting**: Ready for deployment to Hugging Face Hub ## Quick Start ### 1. Install Dependencies ```bash pip install -r requirements.txt ``` ### 2. Start API Server ```bash python api_server.py ``` The server will start at `localhost` Visit `localhost` for interactive API documentation. ### 3. Use the API Client ```bash python api_client.py ``` ## Project Structure ``` ibani-translator/ ├── api_server.py # FastAPI server ├── api_client.py # API client with examples ├── API_USAGE.md # Comprehensive API documentation ├── huggingface_translator.py # Neural translation core ├── rule_based_translator.py # Grammar rules and fallback ├── train_from_ibani_eng.py # Model training script ├── ibani_dict.json # Ibani-English dictionary ├── ibani_eng.json # Training data source ├── ibani_eng.csv # Training data (CSV format) ├── ibani_eng_training_data.json # Formatted training data ├── ibani_model/ # Trained model files ├── requirements.txt # Python dependencies └── README.md # This file ``` ## Usage ### API Translation #### Single Translation ```bash curl -X POST "localhost" \ -H "Content-Type: application/json" \ -d '{"text": "I am eating fish"}' ``` **Response:** ```json { "source": "I am eating fish", "translation": "A nji fịarị", "model": "ibani-translator" } ``` #### Batch Translation ```bash curl -X POST "localhost" \ -H "Content-Type: application …