HachiML/oasst1_for_self-rewarding_EFT_before_judge download history

HachiML/oasst1_for_self-rewarding_EFT_before_judge is a text generation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 28 times (12 in the last 7 days), and 558 times in total. It ranks #399,493 among datasets by monthly downloads.

Dataset Card for Dataset Name kunishou/oasst1-89k-ja OpenAssistant/oasst1 を元に、self-rewardingのEFT(Evaluation Fine-Tuning data)の元データを作成しました。 この後に、学習させたいモデルを使ってLLM-as-a-Judgeを行います。Self-rewardingの論文では最終的に train: 1,630 records, test: 531 records に絞り込んでいます。 Dataset Detail

Open HachiML/oasst1_for_self-rewarding_EFT_before_judge on Hugging Face

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