BART is a denoising autoencoder for pretraining sequence-to-sequence models. It 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 Transformer-based neural machine translation architecture. It uses a standard seq2seq/NMT architecture with a bidirectional encoder (like BERT) and a left-to-right decoder (like GPT). This means the encoder's attention mask is fully visible, like BERT, and the decoder's attention mask is causal, like GPT2.
Source: BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and ComprehensionPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Retrieval | 146 | 13.41% |
Question Answering | 89 | 8.17% |
Language Modelling | 74 | 6.80% |
Text Generation | 66 | 6.06% |
Abstractive Text Summarization | 45 | 4.13% |
Sentence | 41 | 3.76% |
Decoder | 39 | 3.58% |
Text Summarization | 29 | 2.66% |
Information Retrieval | 25 | 2.30% |