Vipplav/tel-input-target

该数据集包含三个主要部分：训练集、测试集和验证集。训练集包含5,554,804个样本，测试集包含685,779个样本，验证集包含617,201个样本。每个样本包含两个序列特征：input_ids和target_ids，这两个特征都是int64类型的序列。数据集的总下载大小为541,878,249字节，总数据集大小为1,096,335,008字节。

The dataset consists of three main parts: a training set, a test set, and a validation set. The training set contains 5,554,804 samples, the test set contains 685,779 samples, and the validation set contains 617,201 samples. Each sample includes two sequence features: input_ids and target_ids, both of which are sequences of int64. The total download size of the dataset is 541,878,249 bytes, and the total dataset size is 1,096,335,008 bytes.

原始信息汇总

数据集概述

数据特征

• input_ids: 序列类型为int64
• target_ids: 序列类型为int64

数据集划分

• train:
  • 字节数: 888032592
  • 样本数: 5554804
• test:
  • 字节数: 109625688
  • 样本数: 685779
• valid:
  • 字节数: 98676728
  • 样本数: 617201

数据集大小

• 下载大小: 541878249 字节
• 总大小: 1096335008 字节

配置

• config_name: default
  • 数据文件路径:
    • train: data/train-*
    • test: data/test-*
    • valid: data/valid-*