kilicai/turkish-sft-v4-jsonl

这是一个土耳其语监督微调（SFT）数据集，版本为v4，以JSONL格式存储。数据集包含108,154个训练示例，每个示例具有id、category、task_type、messages和metadata等特征。messages字段是一个列表，包含角色（role）和内容（content），用于模拟对话交互。metadata包括难度（difficulty）、语言（language，如土耳其语）和质量（quality）等元数据。该数据集由ML Intern工具生成，旨在支持机器学习研究和开发，特别是针对土耳其语的自然语言处理任务。

This is a Turkish Supervised Fine-Tuning (SFT) dataset, version v4, stored in JSONL format. The dataset contains 108,154 training examples, each with features such as id, category, task_type, messages, and metadata. The messages field is a list containing roles and content to simulate dialogue interactions. Metadata includes attributes like difficulty, language (e.g., Turkish), and quality. This dataset was generated by the ML Intern tool and is designed to support machine learning research and development, particularly for Turkish natural language processing tasks.