goelshivam1210/RAPid-Learn

RAPid-Learn : Learn to Recover and Plan Again for Novelties in Open World Environments

This repository contains the code for the RAPidLearn. A framework which integrates planning and learning to solve complex task in non stationary environments for open world learnings. The following code uses the gym-novel-gridworlds environments. The planner used in this case is the Metric FF_v_2.1. For any questions please contact shivam.goel@tufts.edu

Technical Appendix to RAPid-Learn: A Framework for Learning to Recover for Handling Novelties in Open-World Environments. in TechnicalAppendix.pdf

Setup the planner

Download MetricFF.

cd Metric-FF-v2.1
make .

Setup the virtual environment

Use the requirements file to install the dependencies

pip install -r requirements.txt

[Optional] Create a conda environment

conda create --name <YOUR_ENV_NAME> --python=3.7
conda activate <YOUR_ENV_NAME>
pip install -r requirements.txt

Setup the gym_novelgridworld

Install gym_novel_gridworlds.

Switch to branch adaptive_agents

Run the code

To run the RAPidLearn -->

Run
python experiment.py

You can also use arguments.
python experiment.py --TP <trials pre novelty> --TN <trials_post_novelty> --N <novelty_name> --L <learner_name> -E <exploration_mode> -R <render>

novelty_name

1. 'axetobreakeasy'
2. 'axetobreakhard'
3. 'firecraftingtableeasy'
4. 'firecraftingtablehard'
5. 'rubbertree'
6. 'axefirecteasy'
7. 'rubbertreehard'

learner_name

1. 'epsilon-greedy'
2. 'smart-exploration'

exploration_mode

1. 'uniform'
2. 'ucb'

For questions please contact shivam.goel@tufts.edu