Fetching the paper…

Visually-augmented pretrained language models for NLP tasks without images · Around