Horizon: Facebook's Open Source Applied Reinforcement Learning Platform

TL;DR

Horizon is Facebook's open source platform for applying reinforcement learning to large-scale, real-world problems, emphasizing production readiness, data handling, and deployment, with demonstrated success in replacing supervised systems.

Contribution

The paper introduces Horizon, a comprehensive RL platform optimized for industry-scale applications, with features for data processing, distributed training, and deployment, tailored for production environments.

Findings

RL models trained with Horizon outperformed supervised learning systems at Facebook.

Horizon successfully handles large datasets and slow feedback loops in real-world scenarios.

The platform integrates workflows for training, evaluation, and deployment of RL algorithms.

Abstract

In this paper we present Horizon, Facebook's open source applied reinforcement learning (RL) platform. Horizon is an end-to-end platform designed to solve industry applied RL problems where datasets are large (millions to billions of observations), the feedback loop is slow (vs. a simulator), and experiments must be done with care because they don't run in a simulator. Unlike other RL platforms, which are often designed for fast prototyping and experimentation, Horizon is designed with production use cases as top of mind. The platform contains workflows to train popular deep RL algorithms and includes data preprocessing, feature transformation, distributed training, counterfactual policy evaluation, optimized serving, and a model-based data understanding tool. We also showcase and describe real examples where reinforcement learning models trained with Horizon significantly outperformed and replaced supervised learning systems at Facebook.