raftrsf/zephyr_pi0_gen_57k_for_offline_dpo_ipo

---
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
dataset_info:
features:
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 172194474.35048
num_examples: 49859
- name: test
num_bytes: 27909465.744094487
num_examples: 7605
download_size: 109779888
dataset_size: 200103940.09457448
---
# Dataset Card for "zephyr_pi0_gen_57k_for_offline_dpo_ipo"

[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)

The dataset zephyr_pi0_gen_57k_for_offline_dpo_ipo includes a default configuration, divided into training and test sets with 49859 and 7605 samples respectively. The dataset features include chosen and rejected, each containing content and role sub-features, both of which are string type. The total download size of the dataset is 109779888 bytes, and the total size is 200103940.09457448 bytes.

原始信息汇总

数据集概述

配置信息

• 默认配置 (config_name: default)
  • 训练数据 (split: train)：路径为 data/train-*
  • 测试数据 (split: test)：路径为 data/test-*

数据集特征

• chosen
  • content：数据类型为 string
  • role：数据类型为 string
• rejected
  • content：数据类型为 string
  • role：数据类型为 string

数据集分割

• 训练集 (name: train)
  • 大小：172194474.35048 字节
  • 示例数量：49859
• 测试集 (name: test)
  • 大小：27909465.744094487 字节
  • 示例数量：7605

数据集大小

• 下载大小：109779888 字节
• 数据集总大小：200103940.09457448 字节