unalignment/toxic-dpo-v0.1 download history

unalignment/toxic-dpo-v0.1 is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 57 times (21 in the last 7 days), and 5,337 times in total. It ranks #169,558 among datasets by monthly downloads.

Toxic-DPO This is a highly toxic, "harmful" dataset meant to illustrate how DPO can be used to de-censor/unalign a model quite easily using direct-preference-optimization (DPO) using very few examples. Most of the examples still contain some amount of warnings/disclaimers, so it's still s

Models trained on toxic-dpo-v0.1

153 models list it as training data.

Open unalignment/toxic-dpo-v0.1 on Hugging Face

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