autophagycode_metrics_D_metrics_he_unsloth__Qwen3-14B-Base-unsloth-bnb-4bit_lr0.0001_gen5

该数据集包含编程任务相关的结构化数据，主要用于代码质量分析与任务执行评估。数据集包含164个训练样本，总大小18,050字节。每个样本包含15个特征字段：任务索引(task_index)、入口点(entry_point)、是否可执行(is_executable)、是否正确(is_correct)、通过测试数(tests_passed)、失败测试数(tests_failed)、测试运行时间(test_run_time_ms)、错误类型(error_type)等基础字段，以及Halstead复杂度指标（词汇量、长度、体积、难度、工作量）和可维护性指数(maintainability_index)等代码质量度量指标。所有字段均明确标注了数据类型，包括整型(int64)、浮点型(float64)、字符串(string)和布尔型(bool)。数据集仅包含训练集(train)划分。

This dataset contains structured data related to programming tasks, primarily used for code quality analysis and task execution evaluation. The dataset consists of 164 training samples with a total size of 18,050 bytes. Each sample includes 15 feature fields: basic fields such as task_index, entry_point, is_executable, is_correct, tests_passed, tests_failed, test_run_time_ms, error_type, as well as code quality metrics including Halstead complexity metrics (vocabulary size, length, volume, difficulty, effort) and maintainability_index. All fields are clearly annotated with their data types, including int64, float64, string, and bool. The dataset only contains the train split.