데이터셋 탐색

DailyDialog is a high-quality multi-turn open-domain English dialog dataset. It contains 13,118 dialogues split into a training set with 11,118 dialogues and validation and test sets with 1000 dialogues each. On average there are around 8 speaker turns per dialogue with around 15 tokens per turn. Source: http://yanran.li/dailydialog Image Source: https://paperswithcode.com/paper/dailydialog-a-manually-labelled-multi-turn/

Consists of (i) five relevant responses for each context and (ii) five adversarially crafted irrelevant responses for each context. Source: Improving Dialog Evaluation with a Multi-reference Adversarial Dataset and Large Scale Pretraining