AdapterBias: Parameter-efficient Token-dependent Representation Shift for Adapters in NLP Tasks
This addresses storage efficiency for NLP practitioners, but it is incremental as it builds on existing adapter methods.
The paper tackles the problem of large parameter storage in transformer-based pre-trained models by proposing AdapterBias, a simple adapter architecture that adds token-dependent shifts to hidden outputs, reducing trainable parameters dramatically with minimal performance decrease compared to fine-tuned models.
Transformer-based pre-trained models with millions of parameters require large storage. Recent approaches tackle this shortcoming by training adapters, but these approaches still require a relatively large number of parameters. In this study, AdapterBias, a surprisingly simple yet effective adapter architecture, is proposed. AdapterBias adds a token-dependent shift to the hidden output of transformer layers to adapt to downstream tasks with only a vector and a linear layer. Extensive experiments are conducted to demonstrate the effectiveness of AdapterBias. The experiments show that our proposed method can dramatically reduce the trainable parameters compared to the previous works with a minimal decrease in task performances compared with fine-tuned pre-trained models. We further find that AdapterBias automatically learns to assign more significant representation shifts to the tokens related to the task in consideration.