Node-Sampling is an adaptive multi-agent approach to optimize agent sequences.
This repository contains the implementation of the Node-Sampling approach for refinining multiple-choice questions. The main steps are agent definitions and sequence optimization. Optimization can be either performed either by setting a fixed sequence length or by using regularization.
The accompanying Jupyter notebook illustrates the key workflow and examples.
Clone the repository and install dependencies:
git clone https://github.com/UKEIAM/nodesampling
cd Node-Sampling
python -m venv venv
source venv/bin/activate # on Windows use: venv\Scripts\activate
pip install -r requirements.txt