yale-nlp/RefDPO download history

yale-nlp/RefDPO is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 119 times (13 in the last 7 days), and 2,580 times in total. It ranks #132,401 among datasets by monthly downloads.

RefDPO This repo contains the datasets we used in our work "Understanding Reference Policies in Direct Preference Optimization". The datasets are built upon the UltraFeedback dataset, specifically its binarized version ultrafeedback_binarized_cleaned converted from ultrafeedback_binar

Open yale-nlp/RefDPO on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.