APPS (Automated Programming Progress Standard)

APPS 数据集包含从不同的开放访问编码网站（如 Codeforces、Kattis 等）收集的问题。 APPS 基准测试试图通过以不受限制的自然语言提出编码问题并评估解决方案的正确性来反映人类程序员的评估方式。问题的难度范围从入门级到大学竞赛级别，并衡量编码能力以及解决问题的能力。
自动化编程进度标准，缩写为 APPS，总共包含 10,000 个编码问题，有 131,836 个用于检查解决方案的测试用例和 232,444 个由人类编写的真实解决方案。问题可能很复杂，因为问题的平均长度为 293.2 个单词。数据被平均分成训练集和测试集，每个有 5,000 个问题。在测试集中，每个问题都有多个测试用例，平均测试用例数为21.2。每个测试用例都是专门针对相应问题设计的，使我们能够严格评估程序功能。

The APPS dataset contains programming problems collected from various open-access coding websites such as Codeforces, Kattis, and others. The APPS benchmark aims to reflect how human programmers are evaluated by presenting unconstrained natural language-based coding problems and assessing the correctness of their solutions. The problems range in difficulty from entry-level to university competition level, measuring both coding proficiency and problem-solving abilities.

Automated Programming Progress Standard, abbreviated as APPS, includes a total of 10,000 coding problems, with 131,836 test cases for verifying solution correctness and 232,444 real-world solutions written by humans. The problems can be complex, as the average length of a single problem is 293.2 words. The dataset is evenly split into training and test sets, each containing 5,000 problems. In the test set, each problem has multiple test cases, with an average of 21.2 test cases per problem. Each test case is specifically designed for its corresponding problem, enabling rigorous evaluation of program functionality.

搜集汇总

背景与挑战

背景概述

APPS数据集是一个用于评估编程能力的基准测试，包含从开放访问编码网站收集的10,000个编码问题，涵盖入门到竞赛级别难度。该数据集提供了131,836个测试用例和232,444个人类编写的解决方案，数据被平均分为训练集和测试集，每个问题配备多个测试用例以严格评估程序功能。