Hugging Face dataset · #3

Anthropic/hh-rlhf

Anthropic/hh-rlhf is #3 among listed Hugging Face datasets, with 2,449 likes. This repository provides access to two different kinds of data: Human preference data about helpfulness and harmlessness from Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback. These data are meant to train preference (or reward) models…

Dataset: https://huggingface.co/datasets/Anthropic/hh-rlhf
Author: Anthropic (org)
Likes: 2,449
Downloads (30 days): 35,139
License: mit
Size: 100K<n<1M
Tasks: —
Modalities: text
Gated: No
Created: 2022-12-08
Last modified: 2023-05-26
Snapshot: 2026-10-21

Back to Hugging Face datasets · Methodology