Large language models (LLMs) have made significant advancements in code-related tasks, yet many LLMs treat code as simple sequences, neglecting its structured nature. We introduce AST-T5, a novel pretraining paradigm that leverages the Abstract Syntax Tree (AST) for enhanced code generation, transpilation, and understanding. Using dynamic programming, our AST-Aware Segmentation retains code structure, while our AST-Aware Span Corruption objective equips the model to reconstruct various code structures. Unlike other models, AST-T5 avoids intricate program analyses or architectural changes, so it integrates seamlessly with any encoder-decoder Transformer. Evaluations show that AST-T5 consistently outperforms similar-sized LMs across various code-related tasks. Structure-awareness makes AST-T5 particularly powerful in code-to-code tasks, surpassing CodeT5 by 2 points in exact match score for the Bugs2Fix task and by 3 points in exact match score for Java-C# Transpilation in CodeXGLUE. Our code and model are publicly available at https://github.com/gonglinyuan/ast_t5.

翻译：大型语言模型（LLMs）在代码相关任务中取得了显著进展，但许多LLMs将代码视为简单序列，忽略了其结构化特性。我们提出AST-T5，一种新型预训练范式，利用抽象语法树（AST）增强代码生成、转译与理解能力。通过动态规划设计的AST感知分割方法保留了代码结构，而AST感知跨度破坏目标则使模型能够重构多种代码结构。与其他模型不同，AST-T5无需复杂的程序分析或架构调整，因此可无缝集成至任意编码器-解码器Transformer中。评估表明，AST-T5在各类代码相关任务中持续优于同规模语言模型。结构感知特性使AST-T5在代码到代码任务中尤为强大——在CodeXGLUE的Bugs2Fix任务中精确匹配得分超越CodeT5 2分，在Java-C#转译任务中精确匹配得分超越3分。我们的代码与模型已开源至https://github.com/gonglinyuan/ast_t5。