ujgeie/processed_bert_dataset

---
dataset_info:
features:
- name: input_ids
sequence: int32
- name: token_type_ids
sequence: int8
- name: attention_mask
sequence: int8
- name: special_tokens_mask
sequence: int8
splits:
- name: train
num_bytes: 24027285600.0
num_examples: 6674246
download_size: 5887301744
dataset_size: 24027285600.0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

原始信息汇总

数据集信息

特征

• 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
• special_tokens_mask: 序列类型为 int8

分割

• train:
  • 字节数: 24027285600.0
  • 样本数: 6674246

大小

• 下载大小: 5887301744
• 数据集大小: 24027285600.0

配置

• default:
  • 数据文件:
    • 分割: train
    • 路径: data/train-*