Text-to-Text Transfer Transformer
Journal: Journal of Machine Learning Research
Languages: English, French, German, Romanian
Programming languages: Python
Project website: https://github.com/google-research/text-to-text-transfer-transformer
In this paper, we explore the landscape of transfer learning techniques for NLP by introducing a unified framework that converts all text-based language problems into a text-to-text format. Our systematic study compares pre-training objectives, architectures, unlabeled data sets, transfer approaches, and other factors on dozens of language understanding tasks.