OpenAI Generative Pre-Training

In this paper, we explore a semi-supervised approach for language understanding tasks using a combination of unsupervised pre-training and supervised fine-tuning. Our goal is to learn a universal representation that transfers with little adaptation to a wide range of tasks.

Natural language understanding comprises a wide range of diverse tasks such
as textual entailment, question answering, semantic similarity assessment, and
document classification. Although large unlabeled text corpora are abundant,
labeled data for learning these specific tasks is scarce, making it challenging for
discriminatively trained models to perform adequately. We demonstrate that large
gains on these tasks can be realized by generative pre-training of a language model
on a diverse corpus of unlabeled text, followed by discriminative fine-tuning on each
specific task. In contrast to previous approaches, we make use of task-aware input
transformations during fine-tuning to achieve effective transfer while requiring
minimal changes to the model architecture.