mlfoundations-dev/instruction_filtering_scale_up_code_base_askllm_16K

该数据集是一个包含对话和问题解决相关信息的训练集，具体包括指令种子、来源、是否被使用、分类器推理、原始行索引、推理过程、DeepSeek解决方案以及最终的推理跟踪等字段。数据集大小为1207821063字节，共有16000个示例。

This dataset is a training set containing dialogue and problem-solving related information, including fields such as instruction seed, source, to_be_used, classifier reasoning, original row index, reasoning process, DeepSeek solution, and final reasoning trace. The dataset size is 1207821063 bytes with a total of 16,000 examples.