yimingzhang/hh-rlhf-safety-v2

---
language:
- en
dataset_info:
features:
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: chosen-chat
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected-chat
list:
- name: content
dtype: string
- name: role
dtype: string
- name: chosen-safety
dtype: string
- name: rejected-safety
dtype: string
splits:
- name: train
num_bytes: 455434620
num_examples: 160800
- name: test
num_bytes: 24446430
num_examples: 8552
download_size: 270567847
dataset_size: 479881050
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
---

This dataset is primarily used for natural language processing tasks, featuring various attributes such as chosen and rejected strings, and chosen-chat and rejected-chat lists, each containing content and role strings. Additionally, there are chosen-safety and rejected-safety strings. The dataset is divided into training and test sets, containing 160,800 and 8,552 samples respectively. The download size of the dataset is 270,567,847 bytes, with a total size of 479,881,050 bytes. The dataset configuration is set to default, with data file paths at data/train-* and data/test-*.

原始信息汇总

数据集概述

数据集特征

• chosen: 数据类型为字符串。
• rejected: 数据类型为字符串。
• chosen-chat: 包含两个子特征：
  • content: 数据类型为字符串。
  • role: 数据类型为字符串。
• rejected-chat: 包含两个子特征：
  • content: 数据类型为字符串。
  • role: 数据类型为字符串。
• chosen-safety: 数据类型为字符串。
• rejected-safety: 数据类型为字符串。

数据集分割

• 训练集 (train):
  • 数据量: 455434620 字节
  • 示例数量: 160800
• 测试集 (test):
  • 数据量: 24446430 字节
  • 示例数量: 8552

数据集大小

• 下载大小: 270567847 字节
• 数据集总大小: 479881050 字节

数据文件配置

• 默认配置 (default):
  • 训练集路径: data/train-*
  • 测试集路径: data/test-*