gsio-net research

This is an experiment (in-progress) for generating synthetic research in emerging markets.

Prerequisites

  • Anaconda

Quickstart

  1. Clone this repository
  2. Run setup.sh
  3. Follow instructions in Usage

Configuration

The project uses a config.ini file to configure various aspects of the simulation, including:

  • OpenAI API settings
  • Model parameters
  • Simulation settings
  • Logging configuration

Usage

Run the simulation with:

python main.py

This will create a virtual world with predefined agents (Diego, Aisha, Elena, and Nakamura) and start a conversation between them.

Agents

Agents are defined as JSON in the agents/ directory. Each agent file should follow the structure shown in the existing agent files and agents.py

Acknowledgements

Description
Synthetic user research for gsio-net
Readme 36 KiB
Languages
Python 83.8%
Shell 16.2%