Hodfa71/saga-da-delta-dpo-r1 download history

Hodfa71/saga-da-delta-dpo-r1 is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 18 times (3 in the last 7 days), and 137 times in total. It ranks #402,156 among datasets by monthly downloads.

SAGA — Danish Delta-DPO Preference Pairs (Round 1) Preference pairs used to train the Danish Delta-DPO model in the SAGA paper. 7,414 pairs. Each pair has a prompt (first 25–45% of a Danish Wikipedia sentence), a chosen completion and a rejected one. Only pairs with score gap Δ ≥ 0.25 wer

Open Hodfa71/saga-da-delta-dpo-r1 on Hugging Face

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