HachiML/timeseries_QA_multi

该数据集包含多个序列类型的特性，包括input_ids、attention_mask、time_series_values、time_series_input_mask和labels。数据集分为训练集和评估集，训练集包含152014个示例，评估集包含8001个示例。数据集的下载大小为205506653字节，总大小为1556949510.0字节。

This dataset includes multiple sequence-type features such as input_ids, attention_mask, time_series_values, time_series_input_mask, and labels. The dataset is divided into a training set and an evaluation set, with the training set containing 152014 examples and the evaluation set containing 8001 examples. The download size of the dataset is 205506653 bytes, and the total size is 1556949510.0 bytes.

原始信息汇总

数据集概述

数据集特征

• input_ids: 整数序列，类型为 int32
• attention_mask: 整数序列，类型为 int8
• time_series_values: 浮点数序列的序列，类型为 float32
• time_series_input_mask: 浮点数序列，类型为 float32
• labels: 整数序列，类型为 int64

数据集划分

• train:
  • 样本数量: 152014
  • 字节数: 1479099601.994438
• eval:
  • 样本数量: 8001
  • 字节数: 77849908.00556198

数据集大小

• 下载大小: 205506653 字节
• 总大小: 1556949510.0 字节

配置

• config_name: default
  • data_files:
    • train: data/train-*
    • eval: data/eval-*