This is a simple chatbot app using Streamlit. This project is intended to serve the following purposes:
- Utilizes Streamlit to create interactive frontend components, providing a seamless user experience.
- Functions within the application are defined as Pydantic models, leveraging the 'instructor' framework for enhanced structure and validation.
- The app is designed to be easily extendable, with the ability to add new functions and tools with minimal effort.
I know that the code is not perfect, but I am working on it. I am also open to suggestions and contributions. Please bash the code and let me know what you think.
- Multiple GPT Models: Supports various GPT models including GPT-3.5-turbo, GPT-4-0125-preview, and llama2, enabling users to choose their preferred AI assistant.
- Dynamic Response System: Employs a complex system to process user inputs and generate corresponding responses using the selected AI model.
- Custom System Messages: Features a unique chatbot personality, customizable to align with user preferences.
- Streamlit Integration: Built using Streamlit to ensure an engaging and interactive user interface.
- Command Handling: Provides slash commands ('/help', '/clear', '/about') for easy navigation and functionality access.
The application includes various functions to enhance user interaction and utility. Each function is defined as a Pydantic model, ensuring structured and validated input and output. Notable functions include:
- Shell Command Execution: Users can execute safe shell commands directly through the chat interface, enabling a powerful tool for advanced users. [Beware: This function is for demo, executes GPT generated command directly on the host env, I'm using it with safeguards, recommend modifying/using function to be executed in controlled sandbox]
Follow these steps to set up the Draft42 chatbot locally:
- Clone the repository to your local environment.
- Ensure Python 3.8 or newer is installed.
- Install necessary Python dependencies by running
pip install -r requirements.txt. - Launch the Streamlit app with
streamlit run app.py.
After starting the application, select an AI model from the sidebar and interact with the chatbot through the chat input. Use slash commands for additional functionalities:
/help- Shows available commands and usage instructions./clear- Clears the chat history./about- Displays information about the app and its developer.
Some ideas for potential contributions include:
- Adding multiple AI models to simultaneously chat with the user.
- Storing conversations in a lean database(sqlite possibly) for future reference and add it to sidebar.
- Adding more useful functions to chatbot - Plots, Image generation.
For questions, concerns, or suggestions, please contact the project developer Swapnil Patel at:
- Email: [email protected]
- X: @swap357
Draft42 is made available under the MIT License. For more details, refer to the LICENSE file.