mlfoundations-dev/a1_science_kaggle_llm

该数据集包含了多个字段，包括id、prompt、A到E选项、answer答案、instruction_seed指令种子、reasoning推理过程和deepseek_solution解决方案等。数据集分为训练集，共有39997个示例，总文件大小为257,172,995字节。数据集配置为默认配置，训练数据文件以data/train-*的模式存储。

The dataset includes multiple fields such as id, prompt, options A to E, answer, instruction_seed, reasoning, and deepseek_solution. The dataset is split into a training set with a total of 39,997 examples and an overall file size of 257,172,995 bytes. The dataset is configured with a default setting, and the training data files are stored following the pattern data/train-*.