autophagycode_D_metrics_he_Qwen3-8B_lr0.0001_correct_g2

该数据集包含164个训练样本，存储大小为18,987字节，主要记录编程任务相关的执行指标与代码复杂度数据。数据结构包含14个字段：任务ID（task_id）、入口点（entry_point）、可执行状态（is_executable）、正确性标记（is_correct）、通过/失败的测试用例数（tests_passed/tests_failed）、测试运行时间（test_run_time_ms）、错误类型（error_type），以及Halstead复杂度指标（包括词汇量、长度、体积、难度、工作量）和可维护性指数（maintainability_index）。数据仅包含训练集（train split），适用于代码质量分析、缺陷预测或软件度量研究。This dataset comprises 164 training samples with a total storage size of 18,987 bytes, primarily documenting execution metrics and code complexity data related to programming tasks. Its data structure consists of 14 fields: task ID (task_id), entry point (entry_point), executable status (is_executable), correctness flag (is_correct), number of passed/failed test cases (tests_passed/tests_failed), test running time (test_run_time_ms), error type (error_type), as well as Halstead complexity metrics including vocabulary, length, volume, difficulty and effort, and maintainability index (maintainability_index). This dataset only includes the training split, and is applicable to code quality analysis, defect prediction or software metrics research.