five

mlx-community/JOSIE-v2-Instruct-5K

收藏
Hugging Face2026-04-24 更新2026-04-26 收录
下载链接:
https://hf-mirror.com/datasets/mlx-community/JOSIE-v2-Instruct-5K
下载链接
链接失效反馈
官方服务:
资源简介:
JOSIE v2 Instruct 5K是一个高质量的指令跟随数据集,包含5000个对话样本,采用JSONL格式。该数据集用于在Apple Silicon上使用`mlx-lm`或`mlx-lm-lora`微调语言模型。数据集中的每个样本都包含多轮对话,遵循标准的消息格式。JOSIE是一个具有独特个性的AI助手,具有智力深度、干练机智、直接沟通、质量优先、诚实果断和技术精确等特点。数据集涵盖了多个领域,包括高级技术主题、实用编程、科学解释、问题解决、创造性问题和日常查询。数据集生成使用了GPT-5.4-nano模型和OpenAI Batch API,并经过严格的过滤和质量控制。推荐的使用案例包括个性转移、指令跟随、技术写作和Apple Silicon优化。

JOSIE v2 Instruct 5K is a high-quality instruction-following dataset featuring 5,000 conversational samples in JSONL format. It is designed for finetuning language models on Apple Silicon using `mlx-lm` or `mlx-lm-lora`. Each sample contains a multi-turn conversation in the standard messages format. JOSIE is an advanced AI assistant with a distinctive personality combining intellectual rigor, dry wit, genuine helpfulness, direct communication, quality-first approach, honest and decisive answers, and technical precision. The dataset covers diverse domains including advanced technical topics, practical programming, scientific explanations, problem-solving, creative questions, and everyday queries. The dataset was generated using GPT-5.4-nano model and OpenAI Batch API, with rigorous filtering and quality control. Recommended use cases include personality transfer, instruction following, technical writing, and Apple Silicon optimization.
提供机构:
mlx-community
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作