vicgalle/OpenHermesPreferences-roleplay

---
license: other
task_categories:
- text-generation
language:
- en
source_datasets:
- argilla/OpenHermesPreferences
tags:
- dpo
- roleplay
- role-play
- synthetic
- rlaif
- preference
size_categories:
- 1K<n<10K
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
splits:
- name: train
num_bytes: 15514440
num_examples: 3060
download_size: 9727176
dataset_size: 15514440
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

# OpenHermesPreferences-roleplay 🎭


This dataset is a subset from [`argilla/OpenHermesPreferences`](https://hf.co/datasets/argilla/OpenHermesPreferences),
filtered to the following categories: `roleplay, rp, gtkm, greeting`.

To date, it is one of the largest preference datasets specialized towards role-playing applications.

## Usage

The dataset already has the columns `prompt`, `chosen` and `rejected`, so it is trivially compatible with the [DPOTrainer](https://huggingface.co/docs/trl/en/dpo_trainer) from the trl library.

## License

`OpenHermesPreferences-roleplay` inherits the same license as the source dataset [`teknium/OpenHermes-2.5`](https://huggingface.co/datasets/teknium/OpenHermes-2.5) which is currently listed as `other` to account for the varying licenses in each source.

原始信息汇总

数据集概述

基本信息

• 数据集名称: OpenHermesPreferences-roleplay
• 任务类别: 文本生成
• 语言: 英语
• 标签: dpo, roleplay, role-play, synthetic, rlaif, preference
• 大小类别: 1K<n<10K

数据集详情

• 特征:
  • prompt: 字符串
  • chosen: 字符串
  • rejected: 字符串
• 分割:
  • train:
    • 字节数: 15514440
    • 样本数: 3060
• 下载大小: 9727176
• 数据集大小: 15514440

配置

• 配置名称: default
  • 数据文件:
    • split: train
    • path: data/train-*

使用说明

• 数据集包含 prompt, chosen 和 rejected 列，与 DPOTrainer 兼容。

许可证

• 数据集继承自源数据集 teknium/OpenHermes-2.5 的许可证，当前列为 other。