Lyra: A Benchmark for Turducken-Style Code Generation

TL;DR

Lyra introduces a new benchmark dataset for turducken-style code generation, focusing on generating imperative programs with embedded declarative languages from natural language comments, highlighting challenges and potential for real-world applications.

Contribution

This paper defines the first turducken-style code generation task and provides a dataset with annotated programs in Python and embedded SQL, along with baseline evaluations.

Findings

GPT-style models outperform others in accuracy.

AST exact matching accuracy reaches 24-25.5%.

Lyra presents a new challenge for code generation models.

Abstract

Recently, neural techniques have been used to generate source code automatically. While promising for declarative languages, these approaches achieve much poorer performance on datasets for imperative languages. Since a declarative language is typically embedded in an imperative language (i.e., the turducken-style programming) in real-world software development, the promising results on declarative languages can hardly lead to significant reduction of manual software development efforts. In this paper, we define a new code generation task: given a natural language comment, this task aims to generate a program in a base imperative language with an embedded declarative language. To our knowledge, this is the first turducken-style code generation task. For this task, we present Lyra: a dataset in Python with embedded SQL. This dataset contains 2,000 carefully annotated database manipulation programs from real-world projects. Each program is paired with both a Chinese comment and an English comment. In our experiment, we adopted Transformer, BERT-style, and GPT-style models as baselines. In the best setting, the generation performance of GPT-style models is better than others, where the AST exact matching accuracy is 24% and 25.5% when using Chinese and English comments, respectively. Therefore, we believe that Lyra provides a new challenge for code generation. Yet, overcoming this challenge may significantly boost the applicability of code generation techniques for real-world software development.