BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension

Abstract

We present BART, a denoising autoencoder for pretraining sequence-to-sequence models. BART is trained by (1) corrupting text with an arbitrary noising function, and (2) learning a model to reconstruct the original text. It uses a standard Tranformer-based neural machine translation architecture which, despite its simplicity, can be seen as generalizing BERT (due to the bidirectional encoder), GPT (with the left-to-right decoder), and many other more recent pretraining schemes. We evaluate a number of noising approaches, finding the best performance by both randomly shuffling the order of the original sentences and using a novel in-filling scheme, where spans of text are replaced with a single mask token. BART is particularly effective when fine tuned for text generation but also works well for comprehension tasks. It matches the performance of RoBERTa with comparable training resources on GLUE and SQuAD, achieves new state-of-the-art results on a range of abstractive dialogue, question answering, and summarization tasks, with gains of up to 6 ROUGE. BART also provides a 1.1 BLEU increase over a back-translation system for machine translation, with only target language pretraining. We also report ablation experiments that replicate other pretraining schemes within the BART framework, to better measure which factors most influence end-task performance.

Cited by in corpus (17)

• DEPTWEET: A Typology for Social Media Texts to Detect Depression Severities
• Practical Blind Membership Inference Attack via Differential Comparisons
• A Survey on Symbolic Knowledge Distillation of Large Language Models
• LipSound2: Self-Supervised Pre-Training for Lip-to-Speech Reconstruction and Lip Reading
• Flexible, Model-Agnostic Method for Materials Data Extraction from Text Using General Purpose Language Models
• CoderEval: A Benchmark of Pragmatic Code Generation with Generative Pre-trained Models
• Out of the Cage: How Stochastic Parrots Win in Cyber Security Environments
• Hardware-Efficient Photonic Tensor Core: Accelerating Deep Neural Networks with Structured Compression
• On the Effectiveness of Pretrained Models for API Learning
• ChatGPT v.s. Media Bias: A Comparative Study of GPT-3.5 and Fine-tuned Language Models
• Causal-aware Safe Policy Improvement for Task-oriented dialogue
• A Study of Situational Reasoning for Traffic Understanding
• Large Language Models for Combinatorial Optimization: A Systematic Review
• DiffAudit: Auditing Privacy Practices of Online Services for Children and Adolescents
• Infrastructure Ombudsman: Mining Future Failure Concerns from Structural Disaster Response
• HiddenTables & PyQTax: A Cooperative Game and Dataset For TableQA to Ensure Scale and Data Privacy Across a Myriad of Taxonomies
• Divergences between Language Models and Human Brains