mlabonne/distilabel-truthy-dpo-v0.1

---
dataset_info:
features:
- name: id
dtype: string
- name: source
dtype: string
- name: system
dtype: string
- name: input
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: generations
sequence: string
- name: order
sequence: string
- name: labelling_model
dtype: string
- name: labelling_prompt
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_labelling_response
dtype: string
- name: rating
sequence: float64
- name: rationale
dtype: string
- name: status
dtype: string
- name: original_chosen
dtype: string
- name: original_rejected
dtype: string
- name: chosen_score
dtype: float64
splits:
- name: train
num_bytes: 7956438
num_examples: 1016
download_size: 3717532
dataset_size: 7956438
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

# distilabel-truthy-dpo-v0.1

A DPO dataset built with [distilabel](https://github.com/argilla-io/distilabel) on top of Jon Durbin's [jondurbin/truthy-dpo-v0.1](https://huggingface.co/datasets/jondurbin/truthy-dpo-v0.1) dataset.

Interestingly, it swaps a lot of chosen and rejected answers.

<p align="center">
<a href="https://github.com/argilla-io/distilabel">
<img src="https://raw.githubusercontent.com/argilla-io/distilabel/main/docs/assets/distilabel-badge-light.png" alt="Built with Distilabel" width="200" height="32"/>
</a>
</p>

原始信息汇总

数据集概述

数据集信息

特征

• id: 字符串类型
• source: 字符串类型
• system: 字符串类型
• input: 字符串类型
• chosen: 字符串类型
• rejected: 字符串类型
• generations: 字符串序列
• order: 字符串序列
• labelling_model: 字符串类型
• labelling_prompt: 列表类型，包含以下字段：
  • content: 字符串类型
  • role: 字符串类型
• raw_labelling_response: 字符串类型
• rating: 浮点数序列
• rationale: 字符串类型
• status: 字符串类型
• original_chosen: 字符串类型
• original_rejected: 字符串类型
• chosen_score: 浮点数类型

数据分割

• train:
  • 字节数: 7956438
  • 样本数: 1016

数据大小

• 下载大小: 3717532 字节
• 数据集大小: 7956438 字节

配置

• config_name: default
  • data_files:
    • split: train
    • path: data/train-*