README.md

metadata

title: RFT Agent Simulation Engine
emoji: 🧪
colorFrom: blue
colorTo: purple
sdk: gradio
app_file: app.py
pinned: false

RFT Agent Simulation Engine MVP

This project implements a Minimal Viable Product (MVP) of an RFT (Relational Frame Theory) Agent Simulation Engine. It allows for simulating the dynamics of RFT agents, visualizing their behavior, and exporting their states.

Project Structure

• agent.py: Defines the RFTAgent class.
• simulation.py: Defines the RFTSimulation class to manage agent interactions over time.
• visualization.py: Contains functions for plotting simulation results using matplotlib.
• app.py: Implements a Gradio web interface for running simulations and visualizing results.
• utils.py: Provides utility functions for JSON export and seeding random number generators.
• requirements.txt: Lists all Python package dependencies.
• README.md: This file.

How to Run in Colab

1. Open this Colab Notebook: Ensure you have access to the notebook where these files were generated.
2. Run All Cells: Execute all cells in the notebook sequentially. This will create all the necessary .py files, requirements.txt, and README.md in your Colab environment.
3. Install Dependencies: Open a new code cell and run:

   !pip install -r requirements.txt
   

4. Run the Gradio App: To launch the interactive Gradio interface, open a new code cell and run:

   import app
   app.iface.launch(debug=True)
   

   You will get a public URL (or a local URL if running locally) to access the Gradio interface in your browser.

How to Deploy to Hugging Face Spaces

1. Create a New Space: Go to Hugging Face Spaces and create a new Space. Choose Gradio as the SDK.
2. Upload Files: Upload all the generated files (agent.py, simulation.py, visualization.py, app.py, utils.py, requirements.txt, README.md) to your Hugging Face Space repository.
3. Configure app.py: Ensure your app.py has the iface.launch() call at the end (as provided). Hugging Face Spaces automatically detects and runs Gradio apps.
4. Monitor Build Logs: Check the build logs in your Space settings for any dependency installation issues or errors.
5. Access the App: Once successfully built, your Gradio app will be live and accessible via the Space URL.

Example Usage

After running app.iface.launch() in Colab, navigate to the provided URL. You will see a Gradio interface where you can adjust parameters:

• Number of Agents: e.g., 3
• Steps: e.g., 100
• Mutation Rate: e.g., 0.05
• Drift Rate: e.g., 0.02
• Random Seed: e.g., 42

Click the 'Submit' button to run the simulation. The interface will display:

• Plots for individual agent Phi, Tau Effective, and Fitness over time.
• Plots for System Coherence (average Phi) and System Stability (variance of Phi) over time.
• A downloadable JSON file containing the final states of all agents.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference