YDTsai/pretrain_v13_repair7k_102k
收藏Hugging Face2025-02-12 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/YDTsai/pretrain_v13_repair7k_102k
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了tokens、labels、loss_mask和position_ids四个序列化的int64类型字段。数据集分为训练集、验证集和测试集三个部分，其中训练集包含92286个示例，大小为24193697760字节；验证集和测试集各包含5127个示例，大小均为1344094320字节。数据集的总大小为26881886400字节，下载大小为1056332112字节。
The dataset includes four sequence-based fields: tokens, labels, loss_mask, and position_ids, all of which are of type int64. The dataset is split into three parts: training set, validation set, and test set, containing 92,286 examples with a size of 24,193,697,760 bytes for the training set, and 5,127 examples each with a size of 1,344,094,320 bytes for both the validation and test sets. The total size of the dataset is 26,881,886,400 bytes, with a download size of 1,056,332,112 bytes.
提供机构：
YDTsai
