stefanocarrera/autophagycode_D_mercury_Qwen3-8B_lr0.0001_c142_trust_t0.75_g6_run0

该数据集是一个用于任务导向的代码生成或自然语言处理任务的数据集，包含142个训练样本。每个样本具有多个特征：task_id（任务标识符）、entry_point（入口点）、prompt（提示文本）、completion（完成文本）、top_k_progression（top-k进展）和test（测试信息），所有特征均为字符串类型。数据集仅提供train分割，总大小为3791774字节，下载大小为448645字节。数据集未提供明确的中文或英文描述，因此基于结构信息推断其可能用于代码生成、评估或机器学习任务。This dataset is designed for task-oriented code generation or natural language processing tasks, containing 142 training examples. Each example includes multiple features: task_id (task identifier), entry_point (entry point), prompt (prompt text), completion (completion text), top_k_progression (top-k progression), and test (test information), all of which are string types. The dataset only provides a train split, with a total size of 3791774 bytes and a download size of 448645 bytes. No explicit description in English or Chinese is provided in the README, so it is inferred from the structure that it may be used for code generation, evaluation, or machine learning tasks.