cayjobla/wikipedia-pretrain-processed
收藏Hugging Face2023-08-09 更新2024-03-04 收录
下载链接：
https://hf-mirror.com/datasets/cayjobla/wikipedia-pretrain-processed
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: token_type_ids
sequence: int8
- name: attention_mask
sequence: int8
- name: next_sentence_label
dtype: int64
splits:
- name: train
num_bytes: 110639929064.0
num_examples: 35782642
- name: test
num_bytes: 12293328200.0
num_examples: 3975850
download_size: 4973120598
dataset_size: 122933257264.0
---
# Dataset Card for "wikipedia-pretrain-processed"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
提供机构：
cayjobla
原始信息汇总
数据集概述
数据集名称
- 名称: wikipedia-pretrain-processed
数据集特征
- 特征列表:
- input_ids: 序列类型为 int32
- token_type_ids: 序列类型为 int8
- attention_mask: 序列类型为 int8
- next_sentence_label: 数据类型为 int64
数据集分割
- 训练集:
- 样本数量: 35782642
- 存储大小: 110639929064.0 字节
- 测试集:
- 样本数量: 3975850
- 存储大小: 12293328200.0 字节
数据集大小
- 下载大小: 4973120598 字节
- 总数据集大小: 122933257264.0 字节
