Q-bert/wiki-tr-tokenized
收藏Hugging Face2025-11-10 更新2025-11-15 收录
下载链接:
https://hf-mirror.com/datasets/Q-bert/wiki-tr-tokenized
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含训练集,路径为data/train-*。数据集特征包括:id(字符串类型),url(字符串类型),title(字符串类型),input_ids(整型列表),attention_mask(字节数组)。训练集大小为21,867,566,770字节,共有532,601个样本。
The dataset includes a training set with the path data/train-*. The dataset features include: id (string type), url (string type), title (string type), input_ids (list of integers), attention_mask (array of bytes). The training set is 21,867,566,770 bytes in size and contains 532,601 samples.
提供机构:
Q-bert



