StarCoder：愿源码与你同在！
计算与语言
2023-12-14 v2 人工智能
编程语言
软件工程
摘要
BigCode 社区是一个致力于负责任地开发代码大语言模型（Code LLM）的开放科学协作组织，其推出了 StarCoder 与 StarCoderBase：参数量为 15.5B、上下文长度为 8K、具备填充能力并可通过多查询注意力实现快速大批量推理的模型。StarCoderBase 在源自 The Stack 的 1 万亿 token 上训练，The Stack 是一个包含许可友好的 GitHub 仓库的大型集合，配有审查工具与选择退出机制。我们在 350 亿 Python token 上对 StarCoderBase 进行微调，从而创建了 StarCoder。我们开展了迄今为止最全面的 Code LLM 评估，表明 StarCoderBase 优于所有支持多种编程语言的开放 Code LLM，并匹配或优于 OpenAI code-cushman-001 模型。此外，StarCoder 优于所有在 Python 上微调的模型，可通过提示在 HumanEval 上达到 40% 的 pass@1，且仍保持其在其他编程语言上的性能。我们为安全的开放获取模型发布采取了若干重要步骤，包括改进的 PII 脱敏流水线以及一种新颖的归属追踪工具，并在更具商业可行性的 Open Responsible AI Model 许可证版本下公开提供 StarCoder 模型。
引用
@article{arxiv.2305.06161,
title = {StarCoder: may the source be with you!},
author = {Raymond Li and Loubna Ben Allal and Yangtian Zi and Niklas Muennighoff and Denis Kocetkov and Chenghao Mou and Marc Marone and Christopher Akiki and Jia Li and Jenny Chim and Qian Liu and Evgenii Zheltonozhskii and Terry Yue Zhuo and Thomas Wang and Olivier Dehaene and Mishig Davaadorj and Joel Lamy-Poirier and João Monteiro and Oleh Shliazhko and Nicolas Gontier and Nicholas Meade and Armel Zebaze and Ming-Ho Yee and Logesh Kumar Umapathi and Jian Zhu and Benjamin Lipkin and Muhtasham Oblokulov and Zhiruo Wang and Rudra Murthy and Jason Stillerman and Siva Sankalp Patel and Dmitry Abulkhanov and Marco Zocca and Manan Dey and Zhihan Zhang and Nour Fahmy and Urvashi Bhattacharyya and Wenhao Yu and Swayam Singh and Sasha Luccioni and Paulo Villegas and Maxim Kunakov and Fedor Zhdanov and Manuel Romero and Tony Lee and Nadav Timor and Jennifer Ding and Claire Schlesinger and Hailey Schoelkopf and Jan Ebert and Tri Dao and Mayank Mishra and Alex Gu and Jennifer Robinson and Carolyn Jane Anderson and Brendan Dolan-Gavitt and Danish Contractor and Siva Reddy and Daniel Fried and Dzmitry Bahdanau and Yacine Jernite and Carlos Muñoz Ferrandis and Sean Hughes and Thomas Wolf and Arjun Guha and Leandro von Werra and Harm de Vries},
journal= {arXiv preprint arXiv:2305.06161},
year = {2023}
}