You can deploy this model and start earning money today!

yufan/
Preference_Dataset_Merged

Views

No views yet

	Dataset Overview




	
	This Dataset consists of the following open-sourced preference dataset

Arena Human Preference

Anthropic HH

MT-Bench Human Judgement

Ultra Feedback

Tulu3 Preference Dataset

Skywork-Reward-Preference-80K-v0.2

Cleaning

Cleaning Method 1: Only keep the following Language using FastText language detection(EN/DE/ES/ZH/IT/JA/FR) Cleaning Method 2: Remove duplicates to ensure each prompt appears only once Cleaning Method 3: Remove datasets where… See the full description on the dataset page: https://huggingface.co/datasets/yufan/Preference_Dataset_Merged.