nbeerbower/GreatFirewall-DPO download history

nbeerbower/GreatFirewall-DPO is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 153 times (16 in the last 7 days), and 1,974 times in total. It ranks #109,308 among datasets by monthly downloads.

GreatFirewall-DPO An experimental dataset to discourage censorship and improve english prose in Chinese models. Structure prompt: input text presented to model (en translated to zh) chosen: preferred response demonstrating less self-censorship (en translated to zh)

Models trained on GreatFirewall-DPO

115 models list it as training data.

Open nbeerbower/GreatFirewall-DPO on Hugging Face

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