A simple terminal-based AI assistant powered by a lightweight open-source language model (flan-t5-small) using Hugging Face Transformers. This assistant can remember your recent conversation and respond accordingly.
- Lightweight, fast setup (uses
google/flan-t5-small) - Maintains a short conversational history (last 5 turns)
- Runs on CPU β no GPU required
- Simple, readable Python code in a single Jupyter notebook
Personal_AI_Assistant.ipynbβ The main file, contains the full code for running the assistant.
Before running the notebook, make sure you have Python installed (preferably 3.8+). Then install the required Python packages:
pip install transformers torch accelerate -q㪠How to Use
- Open the notebook:
Personal_AI_Assistant.ipynb - Run all cells in order
- Interact with the assistant in the terminal-like interface
Example interaction:
You: where is the white house located?
π€ Assistant: New York City
You: bye
π Assistant: Goodbye!
β Commands
You can end the conversation anytime using:
exitquitbye
π§± Built With
- π€ Transformers
- PyTorch
- flan-t5-small model by Google