Trieste: Efficiently Exploring The Depths of Black-box Functions with TensorFlow

TL;DR

Trieste is an open-source Python toolkit leveraging TensorFlow for scalable Bayesian optimization and active learning, allowing flexible integration of models like Gaussian processes and neural networks for decision-making tasks.

Contribution

The paper introduces Trieste, a modular, extensible library that combines TensorFlow's scalability with Bayesian optimization, supporting custom models and acquisition functions for research and production.

Findings

Supports various TensorFlow-based models like GPflow, GPflux, and Keras.

Provides a flexible, plug-and-play architecture for decision-making loops.

Open-source with comprehensive documentation and testing.

Abstract

We present Trieste, an open-source Python package for Bayesian optimization and active learning benefiting from the scalability and efficiency of TensorFlow. Our library enables the plug-and-play of popular TensorFlow-based models within sequential decision-making loops, e.g. Gaussian processes from GPflow or GPflux, or neural networks from Keras. This modular mindset is central to the package and extends to our acquisition functions and the internal dynamics of the decision-making loop, both of which can be tailored and extended by researchers or engineers when tackling custom use cases. Trieste is a research-friendly and production-ready toolkit backed by a comprehensive test suite, extensive documentation, and available at https://github.com/secondmind-labs/trieste.