five

bennychen/FineTome-100k

收藏
Hugging Face2026-03-25 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/bennychen/FineTome-100k
下载链接
链接失效反馈
官方服务:
资源简介:
--- dataset_info: features: - name: conversations list: - name: from dtype: string - name: value dtype: string - name: source dtype: string - name: score dtype: float64 splits: - name: train num_bytes: 239650960.7474458 num_examples: 100000 download_size: 116531415 dataset_size: 239650960.7474458 configs: - config_name: default data_files: - split: train path: data/train-* --- # FineTome-100k ![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/75I3ffI4XnRlheOQ7kNJ3.jpeg) The FineTome dataset is a subset of [arcee-ai/The-Tome](https://huggingface.co/datasets/arcee-ai/The-Tome) (without arcee-ai/qwen2-72b-magpie-en), re-filtered using [HuggingFaceFW/fineweb-edu-classifier](https://huggingface.co/HuggingFaceFW/fineweb-edu-classifier). It was made for my article ["Fine-tune Llama 3.1 Ultra-Efficiently with Unsloth"](https://huggingface.co/blog/mlabonne/sft-llama3).

数据集信息: 特征字段: - 名称:对话(conversations),为列表类型,其包含子字段: - 字段名:发言方(from),数据类型:字符串(string) - 字段名:内容(value),数据类型:字符串(string) - 名称:来源(source),数据类型:字符串(string) - 名称:评分(score),数据类型:双精度浮点数(float64) 数据集划分: - 划分名称:训练集(train),数据字节数:239650960.7474458,样本数量:100000 下载大小:116531415 数据集总大小:239650960.7474458 配置项: - 配置名称:默认(default),数据文件: - 对应划分:训练集,文件路径:data/train-* # FineTome-100k 数据集 ![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/75I3ffI4XnRlheOQ7kNJ3.jpeg) FineTome数据集是[arcee-ai/The-Tome](https://huggingface.co/datasets/arcee-ai/The-Tome)的子集(已剔除arcee-ai/qwen2-72b-magpie-en子集),并通过[HuggingFaceFW/fineweb-edu-classifier](https://huggingface.co/HuggingFaceFW/fineweb-edu-classifier)完成二次筛选。 本数据集专为我的文章《基于Unsloth实现Llama 3.1的超高效微调》(https://huggingface.co/blog/mlabonne/sft-llama3)制作。
提供机构:
bennychen
二维码
社区交流群
二维码
科研交流群
商业服务