Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/bayesianinstitute/searchgpt
https://github.com/bayesianinstitute/searchgpt
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/bayesianinstitute/searchgpt
- Owner: bayesianinstitute
- Created: 2024-08-20T20:54:40.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2024-08-24T21:12:52.000Z (5 months ago)
- Last Synced: 2024-08-25T21:17:06.234Z (5 months ago)
- Language: Jupyter Notebook
- Size: 67.4 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# **Search LLM**
**Search LLM** is a Python-based project designed to enhance the accuracy and relevance of responses by integrating Large Language Models (LLMs) with real-time web search. This project uses various tools and frameworks, including DuckDuckGo for web search, Google Generative AI for text generation, and FastAPI for efficient API handling.
## **Features**
- **Web Search Integration**: Fetches real-time information from the web using DuckDuckGo.
- **LLM-Based Text Generation**: Generates text responses based on user queries with advanced language models using Gemini (Google Generative AI).
- **Asynchronous API**: Handles requests efficiently with a FastAPI-based backend.
- **Comprehensive Logging**: Monitors and debugs using Loguru for detailed logging.
- **User-Friendly Frontend**: Interact with the system via an intuitive Streamlit web interface.
## **Installation**
1. **Clone the Repository**:
```bash
git clone https://github.com/bayesianinstitute/SearchGPT.git
cd SearchGPT
```
2. **Create a Virtual Environment**:
```bash
python -m venv env
source env/bin/activate # On Windows: `env\Scripts\activate`
```
3. **Install Dependencies**:
```bash
pip install -r requirements.txt
```
4. **Install the Package**:
```bash
python setup.py sdist bdist_wheel
pip install .
```
## **Usage**
1. **Run the FastAPI Server**:
```bash
uvicorn backend.app:app --reload
```
2. **Access the API**:
Open your browser and navigate to `http://127.0.0.1:8000` to interact with the API.
3. **Example Request**:
- **POST /ask/search**: Submit a query to the LLM with optional web search integration.
```bash
curl -X POST "http://127.0.0.1:8000/ask/search" -H "Content-Type: application/json" -d '{"query":"Who won Euro 2024 ?"}'
```
4. **Run the Application (Optional)**:
- **Streamlit UI**:
```bash
cd app
streamlit run UI.py
```
- **Command Line Interface (CLI)**:
```bash
python cli.py
```
## **Project Structure**
- `src/`: Source code for the project.
- `app/`: Contains UI and CLI application code.
- `backend/`: Contains the API service code.
- `requirements.txt`: Lists the required Python packages.
- `setup.py`: Script for building and installing the package.
## **References**
- **[duckduckgo-search](https://pypi.org/project/duckduckgo-search/)**: Python package for web search using DuckDuckGo without API keys.
- **[Loguru](https://loguru.readthedocs.io/)**: A user-friendly logging library for Python.
- **[Google Generative AI](https://developers.generativeai.google/)**: Access Google's Generative AI models for advanced text generation.
- **[Streamlit](https://docs.streamlit.io/)**: Framework for creating custom web apps for data science and machine learning.
- **[FastAPI Documentation](https://fastapi.tiangolo.com/)**: Official documentation for FastAPI, including tutorials and guides.