amuvarma/5k-luna-tune-instruct-6layer-2

该数据集包含用于训练的特征数据，主要包括input_ids、attention_mask和labels三个特征。input_ids表示输入ID的序列，数据类型为int32；attention_mask表示注意力掩码的序列，数据类型为int8；labels表示标签的序列，数据类型为int64。数据集仅包含一个训练集（train），共有2174个样本，总大小为231548392字节。数据集的下载大小为18469601字节。默认配置下，训练集数据文件路径为data/train-*。

This dataset contains feature data for training, mainly including three features: input_ids, attention_mask, and labels. input_ids represents the sequence of input IDs with a data type of int32; attention_mask represents the sequence of attention masks with a data type of int8; labels represents the sequence of labels with a data type of int64. The dataset only includes a training set (train) with 2174 samples and a total size of 231548392 bytes. The download size of the dataset is 18469601 bytes. Under the default configuration, the training set data file path is data/train-*.