Azure99/blossom-v6-sft-stage2

BLOSSOM V6 SFT Stage2是一个高质量、多样化的语言模型微调数据集，旨在为Blossom V6模型的第二阶段SFT训练提供支持。它的目的是进一步增强模型处理更罕见现实世界问题中的复杂指令的能力。数据集主要由中文和英文数据组成，通过三种成本效益高的模型生成不同场景下的响应，并经过基于规则的过滤，以减少重复数据和有毒内容。

BLOSSOM V6 SFT Stage2 is a high-quality, diverse dataset for fine-tuning language models, designed to support the second-stage SFT training of the Blossom V6 model. It aims to enhance the models ability to handle complex instructions in rare real-world scenarios. The dataset consists primarily of Chinese and English data, synthesized using three cost-effective models for generating responses in different scenarios, followed by rule-based filtering to reduce repetitions and toxic content.