Bibliographic Details
| Title: |
基于大模型知识蒸馏的代码摘要自动生成. (Chinese) |
| Alternate Title: |
Code summarization based on large model knowledge distillation. (English) |
| Authors: |
尤 刚, 刘文杰, 李美鹏, 孙立群, 王 炼, 田铁库 |
| Source: |
Command Control & Simulation / Zhihui Kongzhi yu Fangzhen; Aug2025, Vol. 47 Issue 4, p27-33, 7p |
| Subject Terms: |
LANGUAGE models, KNOWLEDGE transfer, GENERATIVE pre-trained transformers, COMPUTER programmers, ABSTRACTION (Computer science), SEMANTICS |
| Abstract (English): |
Code summarization is a short natural language description of source code. Summaries are usually only one sentence long, but they are the primary way for developers to understand code. Recently, products based on large language models (such as ChatGPT) have demonstrated a strong ability to generate these descriptions. However, to use these tools, programmers must send their code to an untrusted third party for processing (for example, through API calls), but this method is unacceptable to many organizations. This paper presents an alternative; we use the example output generated by GPT-3.5 to train an open source model through a process related to knowledge distillation. Enabling small models (with 350 million parameters) to also be comparable to GPT-3.5 in code summarization tasks. [ABSTRACT FROM AUTHOR] |
| Abstract (Chinese): |
代码摘要是对源代码的简短自然语言描述。摘要通常只有一句话的长度, 但却是开发人员了解代码的首要途径。最近, 基于大型语言模型的产品 (如 ChatGPT) 已经展示了生成这些描述的强大能力。不过, 要使用这些工具, 程序员必须将他们的代码发送给不受信任的第三方进行处理 (例如, 通过 API 调用方式),但是许多组织都无法接受这种方式。本文提出了一种替代方案: 使用 GPT-3.5 生成的示例输出, 通过与知识蒸馏相关的过程来训练一个开源模型。使小模型 (3.5 亿参数量) 也能够在代码摘要任务上媲美 GPT-3.5 的效果。 [ABSTRACT FROM AUTHOR] |
|
Copyright of Command Control & Simulation / Zhihui Kongzhi yu Fangzhen is the property of Command Control & Simulation Editorial Office and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Database: |
Complementary Index |
Nájsť tento článok vo Web of Science