Value-Based Deep RL Scales Predictably

TL;DR

This paper demonstrates that value-based off-policy reinforcement learning methods exhibit predictable scaling behavior with data and compute, enabling accurate performance extrapolation and optimal resource allocation.

Contribution

It introduces a method to predict data and compute requirements for desired performance levels and optimize resource allocation in value-based RL.

Findings

Data and compute requirements follow a Pareto frontier controlled by UTD ratio.

Scaling relationships enable performance prediction across different data and compute levels.

Validated approach on multiple algorithms and environments for extrapolating to higher resource levels.

Abstract

Scaling data and compute is critical to the success of modern ML. However, scaling demands predictability: we want methods to not only perform well with more compute or data, but also have their performance be predictable from small-scale runs, without running the large-scale experiment. In this paper, we show that value-based off-policy RL methods are predictable despite community lore regarding their pathological behavior. First, we show that data and compute requirements to attain a given performance level lie on a Pareto frontier, controlled by the updates-to-data (UTD) ratio. By estimating this frontier, we can predict this data requirement when given more compute, and this compute requirement when given more data. Second, we determine the optimal allocation of a total resource budget across data and compute for a given performance and use it to determine hyperparameters that maximize performance for a given budget. Third, this scaling is enabled by first estimating predictable relationships between hyperparameters, which is used to manage effects of overfitting and plasticity loss unique to RL. We validate our approach using three algorithms: SAC, BRO, and PQL on DeepMind Control, OpenAI gym, and IsaacGym, when extrapolating to higher levels of data, compute, budget, or performance.