adamo1139/Poziomka-SFT-v1-mix

这是一个波兰语文本数据集，包含大约200百万个token，用于训练基础模型使用system>user>assistant>user>assistant的格式。数据集包含了不同类型的文本，其中可能包含不适合商业用途的RP/ERP数据。This is a Polish text dataset containing about 20 million tokens, designed for training base models using the system>user>assistant>user>assistant format. The dataset includes various types of texts, some of which may contain RP/ERP data that is not suitable for commercial use.