MortyRAG

Overview

This project, developed by Ant under HermiTech-LLC, implements a Retrieval-Augmented Generation (RAG) model that combines a document retrieval mechanism with a generative language model. The system is designed to produce witty, informative, and contextually appropriate responses, with a particular focus on physics-related content.

---

---

Directory Structure

MortyRAG-main/
├── .github/workflows/           # GitHub Actions workflows
│   └── cd.yml                # CI/CD pipeline configuration
├── custom_t5_rag_local_model_v1.0/
├── data/
│   ├── raw/
│   │   ├── 01_physics_with_wit_and_wisdom.txt
│   │   ├── 02_science_with_a_twist.txt
│   │   ├── 03_sci_fi_and_reality.txt
│   │   ├── 04_black_holes_fact_vs_fiction.txt
│   │   ├── 05_quantum_computing_future.txt
│   │   ├── 06_time_travel_fact_vs_fiction.txt
│   │   ├── 07_gods_of_thunder_mythology.txt
│   │   ├── 08_hidden_wonders_of_earth.txt
│   │   ├── 09_calculus_derivatives.txt
│   │   ├── 10_strange_but_true_history.txt
│   │   ├── 11_rise_of_ai_fiction_vs_reality.txt
│   │   ├── 12_exploring_alien_civilizations.txt
│   │   └── 13_pop_culture_tech_influence.txt
├── docs/
│   ├── api.md
│   ├── controller.md
│   ├── generation.md
│   ├── introduction.md
│   └── retrieval.md
├── LICENSE
├── README.md
├── Screenshot from 2024-08-19 23-18-01.png
├── generator.py         # Custom T5 model class for RAG with local file support
├── main.py              # Entry point script for the Optimization and Query handling GUI
├── rag.py               # Core logic for generating responses using the T5 model
├── requirements.txt     # Required Python packages
├── retriever.py         # Functions for reading and processing different file types
├── Dockerfile           # Dockerfile to build the Docker image for MortyRAG
└── create_shortcut.sh   # Script to create a desktop shortcut to run the Docker container

Installation

Using Docker

MortyRAG is now available as a Docker container, making it easy to deploy and run the application across different environments without needing to install dependencies manually.

1. Pull the Docker Image

First, pull the latest Docker image from the GitHub Container Registry:

docker pull ghcr.io/agreene90/mortyrag:latest

2. Running the Docker Container

To run the MortyRAG application using Docker, execute the following command:

docker run -it --rm -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix ghcr.io/agreene90/mortyrag:latest

This command will launch the MortyRAG Tkinter-based GUI in a Docker container with support for displaying the graphical interface on your local machine.

3. Creating a Desktop Shortcut for Easy Access

You can create a desktop shortcut to run the Dockerized MortyRAG application directly from your desktop.

Creating the Shortcut Script

If you prefer to automate this process, you can run a script that creates the shortcut:

cat << 'EOF' > create_shortcut.sh
#!/bin/bash
set -e

# Pull the latest Docker image
docker pull ghcr.io/agreene90/mortyrag:latest

# Create Desktop Entry for all users
DESKTOP_ENTRY="/usr/share/applications/MortyRAG.desktop"
echo "[Desktop Entry]" > "$DESKTOP_ENTRY"
echo "Version=1.0" >> "$DESKTOP_ENTRY"
echo "Name=MortyRAG" >> "$DESKTOP_ENTRY"
echo "Comment=Start MortyRAG Docker Container" >> "$DESKTOP_ENTRY"
echo "Exec=docker run -it --rm -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix ghcr.io/agreene90/mortyrag:latest" >> "$DESKTOP_ENTRY"
echo "Icon=docker" >> "$DESKTOP_ENTRY"
echo "Terminal=false" >> "$DESKTOP_ENTRY"
echo "Type=Application" >> "$DESKTOP_ENTRY"
echo "Categories=Development;" >> "$DESKTOP_ENTRY"
echo "StartupNotify=true" >> "$DESKTOP_ENTRY"

# Create a symlink to the user's desktop
ln -s "$DESKTOP_ENTRY" "$HOME/Desktop/MortyRAG.desktop"
chmod +x "$HOME/Desktop/MortyRAG.desktop"

echo "MortyRAG has been successfully installed and a shortcut has been added to your desktop."
EOF

Documentation

Detailed documentation for each module can be found in the docs/ directory. Each file provides an in-depth explanation of the module's purpose, usage, and key functions.

License

This project is licensed under the BSD 3-Clause License - see the LICENSE file for details.