CodeTransEngine: Ready-to-use Backend for LLM-based Code Translation

Marcos Macedo, Yuan Tian, Bram Adams

Track: tiny / short paper (up to 4 pages)

Keywords: code translation, benchmarking, tool, system demo, pipeline

TL;DR: CodeTransEngine is a ready-to-use code translation backend that streamlines experiments and lowers the technical barrier for researchers and practitioners by integrating multiple services and algorithms.

Abstract: Code translation, the process of converting a program from one programming language to another, plays a significant role in modernizing legacy systems, ensuring cross-platform compatibility, and improving performance of programs. With the rise of Large Language Models (LLMs), new possibilities have emerged for automating this complex task. However, building effective LLM-based code translation pipelines involves numerous challenges, including prompt engineering, efficient inference, output control, test case validation, and resource optimization. To address these challenges, we introduce CodeTransEngine, a ready-to-use backend that simplifies LLM-based code translation. CodeTransEngine lowers the entry barrier for practitioners and reduces the effort required to leverage the power of LLMs for code translation, enabling them to focus on achieving their translation goals.

Anonymization: This submission has been anonymized for double-blind review via the removal of identifying information such as names, affiliations, and identifying URLs.

Submission Number: 49