Logo Lanfrica

princess-21/Afrolingo

Domain:

natural language processing

Record type:

software
Creator:
pri
Host:
Afrolingo is an AI-powered language platform I am building to help users translate and learn African languages using generative models. It currently focuses on the backend logic and AI integration that generates context-aware translations # Afrolingo – AI Language Platform Prototype Afrolingo is an AI-powered translation project designed to help users translate and learn African languages using deep learning. It uses a TensorFlow sequence model trained on paired text data to translate between African languages and English. ## 🧠 Overview The project explores how artificial intelligence can preserve and promote African languages through intelligent translation models. It uses a **Flask API** for handling translation requests and a **TensorFlow LSTM model** for language generation. Although the current version does not include a frontend interface, the backend logic is fully functional and demonstrates the AI workflow. ## ⚙️ Tech Stack - **Python** - **Flask** - **TensorFlow / Keras** - **scikit-learn** - **NumPy & Pandas** ## 🚀 How to Run 1. Clone the repository: ```bash git clone github.com cd Afrolingo 📝Install dependencies: pip install flask tensorflow scikit-learn pandas numpy Run the Flask app: python app.py Test the translation endpoint: Send a POST request to localhost with JSON data: { "text": "Ndewo", "source": "Igbo", "target": "English" } 📚 How It Works The train_model.py script loads a dataset of paired African and English sentences. Text data is tokenized and converted into padded sequences. A TensorFlow LSTM model is trained to predict the English translation. The Flask app loads the trained model and tokenizer, accepts user text, and returns AI-generated translations. 🌍 Future Plans Add a user interface for text input and translation display. Expand the dataset to cover more African languages. Integrate a multilingual transformer model (e.g., mBART or MarianMT). Deploy the API to the cloud (Render, Hugging Face, or AWS). 👩‍💻 Author Chidera Okeke Ejiroghene GitHub: @princess-21 Medium: @wrlegacy21