seyakin/Health-Chatbot-Nigeria-Language-

Health-Chatbot in Nigeria Languages # Nigerian Health Chatbot ## 1) Requirements * **Python** 3.10+ * **PostgreSQL** running locally (or your own DB connection string) * \~**3 GB disk** free for the NLLB-200 model on first run * An OpenAI API key ```bash # (Recommended) create and activate a virtualenv python -m venv .venv # Windows (PowerShell): . .venv\Scripts\Activate.ps1 # macOS/Linux: source .venv/bin/activate ``` Install dependencies: ```bash pip install --upgrade pip pip install streamlit python-dotenv openai chromadb \ langchain-community langchain-openai \ transformers sentencepiece pypdf pymupdf trafilatura \ sqlalchemy psycopg2-binary requests ``` > If you have a GPU and PyTorch installed, the NLLB model will use it automatically; otherwise it will run on CPU. --- ## 2) Project structure ``` . ├── app.py # Streamlit chat app ├── ingest_health_docs.py # Ingests Nigerian health docs into Chroma ├── .env # Environment variables (create this) ├── data/ │ └── raw/ # Place PDFs/HTML/txt here (or use seed URLs) └── health-knowledge-chroma/ # Auto-created Chroma DB (or your custom dir) ``` --- ## 3) Environment variables (`.env`) Create a file named `.env` in the project root: ```ini # OpenAI OPENAI_API_KEY=sk-xxx # Chroma persistent directory (created automatically) HEALTH_DB_DIR=health-knowledge-chroma # PostgreSQL for chat history (adjust to your setup) DATABASE_URL=postgresql://postgres:YOURPASSWORD@localhost:5432/healthbot # Optional: Bing Web Search key (only used for GPT fallback snippets) # BING_API_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxx ``` --- ## 4) Get Nigerian health documents (optional but recommended) Create folders: **PowerShell (Windows)** ```powershell mkdir data\raw -Force ``` **bash/macOS/Linux** ```bash mkdir -p data/raw ``` Download a starter pack (choose your shell): **Windows PowerShell (use `curl.exe`)** ` …