Louieworth/hh-rlhf-trl-style

---
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
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 73026
num_examples: 50
- name: test
num_bytes: 62643
num_examples: 50
download_size: 87072
dataset_size: 135669
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
---
# TRL's Anthropic HH Dataset

We preprocess the dataset using our standard `prompt, chosen, rejected` format.

## Reproduce this dataset

1. Download the `anthropic_hh.py` from the https://huggingface.co/datasets/Louieworth/hh-rlhf-trl-style/tree/0.1.0.
2. Run `python examples/datasets/anthropic_hh.py --debug --push_to_hub`

原始信息汇总

数据集概述

数据集特征

• chosen
  • content: 数据类型为字符串
  • role: 数据类型为字符串
• rejected
  • content: 数据类型为字符串
  • role: 数据类型为字符串
• prompt: 数据类型为字符串

数据集分割

• train
  • num_bytes: 73026
  • num_examples: 50
• test
  • num_bytes: 62643
  • num_examples: 50

数据集大小

• download_size: 87072
• dataset_size: 135669

配置文件

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