LangProBe: a Language Programs Benchmark

TL;DR

LangProBe is a comprehensive benchmark for evaluating language program architectures and optimization strategies, revealing their impact on quality and cost tradeoffs in AI systems.

Contribution

It introduces the first large-scale benchmark for language programs, enabling systematic study of architecture and optimizer effects on performance and efficiency.

Findings

Optimized language programs improve cost-quality tradeoffs.

Human judgment remains crucial for selecting optimal compositions.

Benchmark facilitates systematic evaluation of language program strategies.

Abstract

Composing language models (LMs) into multi-step language programs and automatically optimizing their modular prompts is now a mainstream paradigm for building AI systems, but the tradeoffs in this space have only scarcely been studied before. We introduce LangProBe, the first large-scale benchmark for evaluating the architectures and optimization strategies for language programs, with over 2000 combinations of tasks, architectures, optimizers, and choices of LMs. Using LangProBe, we are the first to study the impact of program architectures and optimizers (and their compositions together and with different models) on tradeoffs of quality and cost. We find that optimized language programs offer strong cost--quality Pareto improvement over raw calls to models, but simultaneously demonstrate that human judgment (or empirical decisions) about which compositions to pursue is still necessary for best performance. We will open source the code and evaluation data for LangProBe.