mlfoundations-dev/instruction_filtering_scale_up_code_base_gemini_length_8K

这是一个包含对话和问题解决相关信息的文本数据集。数据集的特征包括指令种子、来源、Gemini模型响应、原始行索引、长度、推理过程、DeepSeek解决方案和最终的推理轨迹。此外，还包含了对话信息，其中有对话的发起者和内容。数据集分为训练集，共有8000个示例，大小约为595MB。数据集的配置信息中，默认配置下训练集的数据文件路径以data/train-开头。

This is a text dataset containing dialogues and problem-solving related information. The dataset features include instruction seed, source, Gemini model response, original row index, length, reasoning process, DeepSeek solution, and the final reasoning trace. Additionally, it includes conversation information with the initiator and content of the conversation. The dataset is split into a training set with a total of 8000 examples, approximately 595MB in size. In the default configuration, the data file path for the training set starts with data/train-.