The DeBERTa model was proposed in DeBERTa: Decoding-enhanced BERT with Disentangled Attention by Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen It is based on Google's BERT model released in 2018 and Facebook's RoBERTa model released in 2019.
It builds on RoBERTa with disentangled attention and enhanced mask decoder training with half of the data used in RoBERTa.
The abstract from the paper is the following:
Recent progress in pre-trained neural language models has significantly improved the performance of many natural language processing (NLP) tasks. In this paper we propose a new model architecture DeBERTa (Decoding-enhanced BERT with disentangled attention) that improves the BERT and RoBERTa models using two novel techniques. The first is the disentangled attention mechanism, where each word is represented using two vectors that encode its content and position, respectively, and the attention weights among words are computed using disentangled matrices on their contents and relative positions. Second, an enhanced mask decoder is used to replace the output softmax layer to predict the masked tokens for model pretraining. We show that these two techniques significantly improve the efficiency of model pretraining and performance of downstream tasks. Compared to RoBERTa-Large, a DeBERTa model trained on half of the training data performs consistently better on a wide range of NLP tasks, achieving improvements on MNLI by +0.9% (90.2% vs. 91.1%), on SQuAD v2.0 by +2.3% (88.4% vs. 90.7%) and RACE by +3.6% (83.2% vs. 86.8%). The DeBERTa code and pre-trained models will be made publicly available at https://github.com/microsoft/DeBERTa.
The following information is visible directly on the original implementation repository. DeBERTa v2 is the second version of the DeBERTa model. It includes the 1.5B model used for the SuperGLUE single-model submission and achieving 89.9, versus human baseline 89.8. You can find more details about this submission in the authors' blog
New in v2:
- Vocabulary In v2 the tokenizer is changed to use a new vocabulary of size 128K built from the training data. Instead of a GPT2-based tokenizer, the tokenizer is now sentencepiece-based tokenizer.
- nGiE(nGram Induced Input Encoding) The DeBERTa-v2 model uses an additional convolution layer aside with the first transformer layer to better learn the local dependency of input tokens.
- Sharing position projection matrix with content projection matrix in attention layer Based on previous experiments, this can save parameters without affecting the performance.
- Apply bucket to encode relative positions The DeBERTa-v2 model uses log bucket to encode relative positions similar to T5.
- 900M model & 1.5B model Two additional model sizes are available: 900M and 1.5B, which significantly improves the performance of downstream tasks.
This model was contributed by DeBERTa. This model TF 2.0 implementation was contributed by kamalkraj. The original code can be found here.
- Text classification task guide
- Token classification task guide
- Question answering task guide
- Masked language modeling task guide
- Multiple choice task guide
[[autodoc]] DebertaV2Config
[[autodoc]] DebertaV2Tokenizer - build_inputs_with_special_tokens - get_special_tokens_mask - create_token_type_ids_from_sequences - save_vocabulary
[[autodoc]] DebertaV2TokenizerFast - build_inputs_with_special_tokens - create_token_type_ids_from_sequences
[[autodoc]] DebertaV2Model - forward
[[autodoc]] DebertaV2PreTrainedModel - forward
[[autodoc]] DebertaV2ForMaskedLM - forward
[[autodoc]] DebertaV2ForSequenceClassification - forward
[[autodoc]] DebertaV2ForTokenClassification - forward
[[autodoc]] DebertaV2ForQuestionAnswering - forward
[[autodoc]] DebertaV2ForMultipleChoice - forward
[[autodoc]] TFDebertaV2Model - call
[[autodoc]] TFDebertaV2PreTrainedModel - call
[[autodoc]] TFDebertaV2ForMaskedLM - call
[[autodoc]] TFDebertaV2ForSequenceClassification - call
[[autodoc]] TFDebertaV2ForTokenClassification - call
[[autodoc]] TFDebertaV2ForQuestionAnswering - call
[[autodoc]] TFDebertaV2ForMultipleChoice - call