Deep Reinforcement Learning with Task-Adaptive Retrieval via Hypernetwork

This repository is for our paper 'Deep Reinforcement Learning with Task-Adaptive Retrieval via Hypernetwork'(https://arxiv.org/abs/2306.10698).

Model

Installation

1. Clone this repository.

2. Install requirements.

cd task-hypernet
conda create -n minigrid python=3.9
pip install -r requirements.txt

Example of use

Train an agent.

python3 -m scripts.train --algo ppo --model model --save-interval 10 --frames 80000