Iterated Population Based Training with Task-Agnostic Restarts
This work addresses a specific bottleneck in hyperparameter optimization for machine learning practitioners, offering an incremental improvement over existing PBT methods.
The paper tackled the problem of efficiently setting the number of steps between hyperparameter updates in Population Based Training (PBT) variants by introducing Iterated Population Based Training (IPBT), which automatically adjusts this hyperparameter via task-agnostic restarts and time-varying Bayesian optimization, resulting in matching or outperforming 5 previous PBT variants and other HPO algorithms on 8 image classification and reinforcement learning tasks without budget increases.
Hyperparameter Optimization (HPO) can lift the burden of tuning hyperparameters (HPs) of neural networks. HPO algorithms from the Population Based Training (PBT) family are efficient thanks to dynamically adjusting HPs every few steps of the weight optimization. Recent results indicate that the number of steps between HP updates is an important meta-HP of all PBT variants that can substantially affect their performance. Yet, no method or intuition is available for efficiently setting its value. We introduce Iterated Population Based Training (IPBT), a novel PBT variant that automatically adjusts this HP via restarts that reuse weight information in a task-agnostic way and leverage time-varying Bayesian optimization to reinitialize HPs. Evaluation on 8 image classification and reinforcement learning tasks shows that, on average, our algorithm matches or outperforms 5 previous PBT variants and other HPO algorithms (random search, ASHA, SMAC3), without requiring a budget increase or any changes to its HPs. The source code is available at https://github.com/AwesomeLemon/IPBT.