1 dataset found

Groups: Vision-Language Pre-training Formats: JSON

Filter Results
  • BookCorpus

    The dataset used in this paper for unsupervised sentence representation learning, consisting of paragraphs from unlabeled text.