norasalem88/rag_assistant_devops

1---2title: DevOps RAG Assistant3emoji: 🤖4colorFrom: blue5colorTo: green6sdk: streamlit7sdk_version: 1.44.18app_file: app.py9pinned: false10---11 12# DevOps RAG Assistant13 14Upload your DevOps PDFs and ask questions. Powered by Hugging Face.15 16 17## Features18- **Local Execution:** Uses ChromaDB for vector storage and Ollama (qwen2.5:7b) for generation.19- **English Only:** Focused specifically on DevOps documentation in English.20- **Minimal UI:** Streamlit interface for uploading PDFs and asking questions.21 22## Setup231. **Ensure Ollama is installed** and the `qwen2.5:7b` model is pulled:24   ```bash25   ollama pull qwen2.5:7b26   ```272. **Setup Python environment:**28   ```bash29   python -m venv venv30   # Windows31   .\venv\Scripts\activate32   # Linux/Mac33   source venv/bin/activate34   ```353. **Install dependencies:**36   ```bash37   pip install -r requirements.txt38   ```39 40## Running the Application41To run the Streamlit UI:42```bash43streamlit run app.py44```45 46Then open your browser to `http://localhost:8501`.47You can upload DevOps PDFs (like Kubernetes & Docker documentation) via the sidebar.48 49## Project Structure50- `data/`: Contains uploaded PDF documents.51- `src/`: 52  - `ingestion.py`: Handles loading PDFs, chunking, and embedding creation.53  - `retrieval.py`: Setup for ChromaDB retriever.54  - `generation.py`: LangChain RAG pipeline connecting ChromaDB and Ollama.55  - `utils.py`: Common helper functions (logging, embeddings).56- `vectorstore/`: Persisted ChromaDB embeddings.57- `app.py`: Streamlit frontend.58