Aurumdev952/kinyarwanda-assistant

# Mini Kinyarwanda Voice Assistant This project implements a basic voice assistant pipeline for the Kinyarwanda language, simulating the core components of an intelligent robot's voice interaction capabilities: Automatic Speech Recognition (ASR), Natural Language Processing (NLP) for understanding, and Text-to-Speech (TTS) for responding. It was developed as an assignment for the Intelligent Robotics course (Assigned: April 23, 2025, Due: April 30, 2025). ## Features * **ASR:** Converts Kinyarwanda speech to text using a fine-tuned Whisper model (`benax-rw/KinyaWhisper`). * **NLP:** Understands simple Kinyarwanda questions by matching transcriptions against a predefined dictionary of Q&A pairs. * **TTS:** Generates spoken Kinyarwanda responses using the Coqui TTS library. * **UI:** Provides a simple web interface using Gradio to interact with the voice assistant via microphone ## Technologies Used * Python 3 * Hugging Face `transformers` (for KinyaWhisper ASR) * `pytorch` (backend for transformers) * `gTTS` (for Kinyarwanda TTS) * `gradio` (for the web UI) * `librosa`, `soundfile`, `torchaudio` (for audio handling) * `conda` (for environment management) ## Setup 1. **Clone the repository:** ```bash git clone github.com assistant cd assistant ``` 2. **Create and activate the Conda environment:** Ensure you have Conda installed. ```bash conda env create -f transformers-audio.yml conda activate transformers-audio ``` 3. **Verify Dependencies:** The `transformers-audio.yml` file should handle most dependencies. If you encounter issues, ensure `pytorch` (CPU version is specified) and the other listed libraries are installed correctly within the `transformers-audio` environment. ## Running the Application 1. Activate the conda environment: ```bash conda activate transformers-audio ``` 2. Run the Gradio application script: ```bash python app.py ``` This will start a local web server, and t …