The Well: a Large-Scale Collection of Diverse Physics Simulations for Machine Learning

TL;DR

The paper introduces 'The Well,' a comprehensive collection of large-scale, diverse physics simulation datasets designed to advance machine learning surrogate models and provide a broad benchmark for evaluating new approaches.

Contribution

It provides a large, diverse, and unified dataset collection with a PyTorch interface, addressing limitations of existing small, domain-specific datasets in physics-based machine learning.

Findings

Demonstrated the utility of the dataset with baseline models.

Highlighted new challenges in modeling complex physical dynamics.

Provided open access to data and tools for the community.

Abstract

Machine learning based surrogate models offer researchers powerful tools for accelerating simulation-based workflows. However, as standard datasets in this space often cover small classes of physical behavior, it can be difficult to evaluate the efficacy of new approaches. To address this gap, we introduce the Well: a large-scale collection of datasets containing numerical simulations of a wide variety of spatiotemporal physical systems. The Well draws from domain experts and numerical software developers to provide 15TB of data across 16 datasets covering diverse domains such as biological systems, fluid dynamics, acoustic scattering, as well as magneto-hydrodynamic simulations of extra-galactic fluids or supernova explosions. These datasets can be used individually or as part of a broader benchmark suite. To facilitate usage of the Well, we provide a unified PyTorch interface for training and evaluating models. We demonstrate the function of this library by introducing example baselines that highlight the new challenges posed by the complex dynamics of the Well. The code and data is available at https://github.com/PolymathicAI/the_well.