A conversational AI chat desktop application built with Python, Tkinter, and Hugging Face Transformers. Eva remembers recent conversations, responds naturally, and simulates a personalized AI companion.
-
Chat with Eva in a user-friendly GUI
-
Memory of the last few conversation turns
-
Personalized responses based on a user profile
-
Typing indicator while Eva is generating a response
-
Commands:
/clear— clear chat window/reset— reset conversation memory
-
Color-coded messages for human and Eva
-
Threaded AI response generation for smooth UI
- Python 3
- Tkinter – GUI framework
- PyTorch – for model inference
- Transformers (Hugging Face) – AI model and tokenizer
- JSON – storing user profile and conversation memory
- Threading – responsive UI during AI computation
Clone the repository:
git clone https://github.com/ErkanSoftwareDeveloper/eva-assistant.git
cd eva-assistantInstall dependencies:
pip install -r requirements.txtMake sure you have PyTorch installed with a compatible version for your system.
Run the application:
python Eva.py- Type a message and press Enter to send
- Use
/clearto clear the chat window - Use
/resetto reset Eva’s memory - Chat with Eva and watch her respond naturally
eva-assistant/
├─ Eva.py # Main application file
├─ profile.json # User profile for personalized responses
├─ .gitignore # Git ignored files
├─ README.md # Project documentation
└─ requirements.txt # Project dependencies
- Voice input/output
- Conversation history saving/loading
- Multi-profile support
- Dark/light theme UI
- Packaging as executable (.exe)
This project is intended for educational and personal use.
