mlfoundations-dev/a1_code_primeintellect_stack_exchange

该数据集包含了源文本(source)、任务类型(task_type)、指令种子(instruction_seed)、黄金标准解决方案(gold_standard_solution)、验证信息(verification_info)、元数据(metadata)、问题ID(problem_id)、推理过程(reasoning)、deepseek解决方案(deepseek_solution)等字段。数据集分为训练集(train)，包含40000个示例，总大小为395,638,987字节。

The dataset includes fields such as source text (source), task type (task_type), instruction seed (instruction_seed), gold standard solution (gold_standard_solution), verification information (verification_info), metadata (metadata), problem ID (problem_id), reasoning process (reasoning), and deepseek solution (deepseek_solution). The dataset is split into a training set (train) containing 40,000 examples, with a total size of 395,638,987 bytes.