# Metaoptimization on a Distributed System for Deep Reinforcement Learning

## TL;DR

This paper introduces HyperTrick, a new metaoptimization algorithm designed to efficiently tune hyperparameters in distributed deep reinforcement learning, leading to more stable training and better resource utilization.

## Contribution

The paper presents HyperTrick, a novel, simpler metaoptimization method that improves hyperparameter tuning in distributed deep reinforcement learning systems.

## Key findings

- HyperTrick effectively tunes hyperparameters for deep RL agents.
- It achieves similar policy performance with more efficient resource use.
- HyperTrick outperforms state-of-the-art metaoptimization algorithms in distributed settings.

## Abstract

Training intelligent agents through reinforcement learning is a notoriously unstable procedure. Massive parallelization on GPUs and distributed systems has been exploited to generate a large amount of training experiences and consequently reduce instabilities, but the success of training remains strongly influenced by the choice of the hyperparameters. To overcome this issue, we introduce HyperTrick, a new metaoptimization algorithm, and show its effective application to tune hyperparameters in the case of deep reinforcement learning, while learning to play different Atari games on a distributed system. Our analysis provides evidence of the interaction between the identification of the optimal hyperparameters and the learned policy, that is typical of the case of metaoptimization for deep reinforcement learning. When compared with state-of-the-art metaoptimization algorithms, HyperTrick is characterized by a simpler implementation and it allows learning similar policies, while making a more effective use of the computational resources in a distributed system.