chemige/IF_multi_constraints_upto5_SFT

该监督数据集源自allenai/IF_multi_constraints_upto5指令遵循基准。响应由google/gemma-4-31B-it生成，并经过筛选仅保留高质量示例。最终数据集仅包含最低宽松分数1.0、最低严格分数0.8和最大响应长度1024个标记的样本。该数据集旨在用于多约束指令遵循任务的监督微调，强调对请求约束的严格遵守和简洁输出。

This supervised dataset is derived from the `allenai/IF_multi_constraints_upto5` instruction-following benchmark. Responses were generated with `google/gemma-4-31B-it` and then filtered to retain only high-quality examples. The final dataset includes only samples with a minimum loose score of `1.0`, a minimum strict score of `0.8`, and a maximum response length of `1024` tokens. It is intended for supervised fine-tuning on multi-constraint instruction-following tasks, with an emphasis on strong adherence to requested constraints and concise outputs.