autophagycode_D_metrics_he_Qwen3-14B_lr0.0001_scm_g10
收藏资源简介:
该数据集包含164个编程任务相关的执行记录与代码质量分析数据。每个样本记录包含16个结构化字段:任务标识符(task_id)、可执行状态(is_executable)、正确性标记(is_correct)、通过/失败的测试案例计数(tests_passed/tests_failed)、错误类型(error_type)等执行信息,以及Halstead复杂度指标(halstead_vocabulary/length/volume/difficulty/effort)、可维护性指数(maintainability_index)、定义函数数量(n_func_defined)等静态分析指标。数据集仅包含训练集(train),总大小18.4KB,适用于代码质量评估、程序错误检测或自动化测试生成等任务。
This dataset contains 164 execution records and code quality analysis data related to programming tasks. Each sample includes 16 structured fields: task identifier (task_id), executable status (is_executable), correctness flag (is_correct), pass/fail test case counts (tests_passed/tests_failed), error type (error_type) and other execution information, as well as static analysis metrics such as Halstead complexity metrics (halstead_vocabulary/length/volume/difficulty/effort), maintainability index (maintainability_index), number of defined functions (n_func_defined) and others. The dataset only includes the training set (train), with a total size of 18.4 KB, and is applicable to tasks such as code quality assessment, program bug detection or automated test generation.
数据集概述
基本信息
- 数据集名称: autophagycode_D_metrics_he_Qwen3-14B_lr0.0001_scm_g10
- 来源地址: https://huggingface.co/datasets/stefanocarrera/autophagycode_D_metrics_he_Qwen3-14B_lr0.0001_scm_g10
- 数据量: 164 个样本
- 数据集大小: 18422 字节
- 下载大小: 16804 字节
数据结构
数据集包含一个名为 train 的分割,包含以下特征字段:
task_id: 字符串类型,任务标识符。entry_point: 字符串类型,入口点。is_executable: 布尔类型,是否可执行。is_correct: 布尔类型,是否正确。tests_passed: 整型,通过的测试数。tests_failed: 整型,失败的测试数。test_run_time_ms: 空值类型,测试运行时间(毫秒)。error_type: 字符串类型,错误类型。halstead_vocabulary: 整型,Halstead 词汇量。halstead_length: 整型,Halstead 长度。halstead_volume: 浮点型,Halstead 体积。halstead_difficulty: 浮点型,Halstead 难度。halstead_effort: 浮点型,Halstead 工作量。maintainability_index: 浮点型,可维护性指数。n_func_defined: 整型,定义的函数数量。entry_point_repeated: 布尔类型,入口点是否重复。
配置信息
- 配置名称: default
- 数据文件:
train分割的数据文件路径为data/train-*。




