Zlovoblachko/REALEC_GED-transformer-first_level
收藏Hugging Face2025-04-06 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/Zlovoblachko/REALEC_GED-transformer-first_level
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了输入ID、token类型ID、注意力掩码和标签四个特征。输入ID和标签是序列化的int32和float64类型数据，而token类型ID和注意力掩码是序列化的int8类型数据。数据集分为训练集、验证集、测试集和重合集四部分，分别包含了61603、7700、7701和8427个样本。数据集的总大小为31,733,330字节，下载大小为4,812,094字节。
The dataset includes four features: input IDs, token type IDs, attention masks, and labels. Input IDs and labels are serialized int32 and float64 data types, respectively, while token type IDs and attention masks are serialized int8 data types. The dataset is divided into four parts: training set, validation set, test set, and overlapping set, containing 61,603, 7,700, 7,701, and 8,427 samples, respectively. The total size of the dataset is 31,733,330 bytes, and the download size is 4,812,094 bytes.
提供机构：
Zlovoblachko
