codedp-bench-repo-patch
收藏Hugging Face2026-03-16 更新2026-03-20 收录
下载链接:
https://huggingface.co/datasets/melihcatal/codedp-bench-repo-patch
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含代码相关的结构化信息,涵盖18个特征字段,主要包括任务ID(task_id)、代码库来源(repo)、文件路径(file_path)、函数名称(function_name)、完整限定名(qualified_name)、函数类型(function_type)、类名(class_name)、提示文本(prompt)、函数签名(signature)、文档字符串(docstring)、规范解决方案(canonical_solution)、完整函数(full_function)、测试用例(tests)、环境配置(setup)、元数据(metadata)、验证信息(validation)、原始任务ID(original_task_id)和污染标签(contamination_label)。数据集仅包含训练集(train),共36个样本,总大小约3.2MB。从字段命名推断,该数据集可能适用于代码生成、文档自动化或软件测试等编程相关任务。
This dataset contains code-related structured information, encompassing 18 feature fields, specifically task_id, repo, file_path, function_name, qualified_name, function_type, class_name, prompt, signature, docstring, canonical_solution, full_function, tests, setup, metadata, validation, original_task_id, and contamination_label. The dataset only includes the training split (train), with a total of 36 samples and an approximate total size of 3.2 MB. Based on the naming conventions of its fields, this dataset can be applied to programming-related tasks such as code generation, documentation automation, and software testing.
创建时间:
2026-03-05



