Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

【论文分享-1212】Evaluating Large Language Models Trained on Code #50

Open
LPY22 opened this issue Dec 12, 2023 · 0 comments
Open
Labels
研究分享:推荐 好论文推荐

Comments

@LPY22
Copy link

LPY22 commented Dec 12, 2023

Title

Evaluating Large Language Models Trained on Code

Link

https://arxiv.org/pdf/2107.03374.pdf

Year

2021

Conference or Journal

Preprint

Rank

No response

Keywords

GPT Code generation

Abstract

本文介绍了 Codex,这是一个在 GitHub 的公开可用代码上微调的 GPT 语言模型,并研究了它的 Python 代码编写能力。Codex 的另一种生产版本支持了 GitHub Copilot 。在 HumanEval 上,本文发布了一个新的评估集来衡量代码生成功能正确性,我们的模型解决了数据集 28.8% 的问题,而 GPT-3 解决了 0%,GPT-J 解决了 11.4%。此外,我们发现从模型中重复采样是产生解决方案的有效策略。使用这种方法,我们解决了每个问题有 100 个样本的问题的 70.2%。仔细调查我们的模型揭示了其局限性,包括文档描述的长链操作以及对变量的绑定操作。最后,我们讨论了部署强大的代码生成技术的潜在更广泛的影响。

@LPY22 LPY22 added the 研究分享:推荐 好论文推荐 label Dec 12, 2023
@LPY22 LPY22 changed the title [Paper Recommendation] 【论文分享-1212】Evaluating Large Language Models Trained on Code Dec 12, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
研究分享:推荐 好论文推荐
Projects
None yet
Development

No branches or pull requests

1 participant