ekolasky/NQLongAnsForLSGSentClassWSTok
收藏Hugging Face2024-01-09 更新2024-03-04 收录
下载链接：
https://hf-mirror.com/datasets/ekolasky/NQLongAnsForLSGSentClassWSTok
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
- name: global_attention_mask
sequence: int64
- name: dataset_index
dtype: int64
splits:
- name: train
num_bytes: 5096195263
num_examples: 58195
download_size: 552583139
dataset_size: 5096195263
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---
The dataset includes multiple features such as input_ids, attention_mask, labels, global_attention_mask, and dataset_index, each corresponding to different data types. The dataset is divided into a training set (train) with 58195 examples and a total size of 5096195263 bytes. The download size of the dataset is 552583139 bytes. Additionally, the dataset has a default configuration (default) with training data file paths at data/train-*.
提供机构：
ekolasky
原始信息汇总
数据集概述
数据特征
- input_ids: 序列类型，数据类型为int32。
- attention_mask: 序列类型，数据类型为int8。
- labels: 序列类型，数据类型为int64。
- global_attention_mask: 序列类型，数据类型为int64。
- dataset_index: 数据类型为int64。
数据分割
- train: 包含58195个样本，总字节数为5096195263。
数据集大小
- 下载大小: 552583139字节。
- 数据集大小: 5096195263字节。
配置信息
- config_name: default
- data_files:
- split: train
- path: data/train-*
- data_files:
