Investigating Efficiently Extending Transformers for Long Input Summarization
While large pretrained Transformer models have proven highly capable at tackling natural language tasks, handling long sequence inputs continues to be a significant challenge. One such task is long input summarization, where inputs are longer than the maximum input context of most pretrained models. Through an extensive set of experiments, we investigate what model architectural changes and pretraining paradigms can most efficiently adapt a pretrained Transformer for long input summarization. We find that a staggered, block-local Transformer with global encoder tokens strikes a good balance of performance and efficiency, and that an additional pretraining phase on long sequences meaningfully improves downstream summarization performance. Based on our findings, we introduce PEGASUS-X, an extension of the PEGASUS model with additional long input pretraining to handle inputs of up to 16K tokens. PEGASUS-X achieves strong performance on long input summarization tasks comparable with much larger models while adding few additional parameters and not requiring model parallelism to train.
Code (2)
Tasks
16kLong-range modelingText SummarizationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Functional Interpolation for Relative Positions Improves Long Context Transformers
Preventing the performance decay of Transformers on inputs longer than those used for training has been an important challenge in extending the context length of these models. Though the Transformer architecture has fund…
Language ModelingLanguage ModellingPositionUnlimiformer: Long-Range Transformers with Unlimited Length Input
Since the proposal of transformers, these models have been limited to bounded input lengths, because of their need to attend to every token in the input. In this work, we propose Unlimiformer: a general approach that wra…
Book summarizationCPUDecoderGPUInvestigating Recurrent Transformers with Dynamic Halt
In this paper, we comprehensively study the inductive biases of two major approaches to augmenting Transformers with a recurrent mechanism: (1) the approach of incorporating a depth-wise recurrence similar to Universal T…
DiagnosticLanguage ModelingLanguage ModellingListOpsCurved Representation Space of Vision Transformers
Neural networks with self-attention (a.k.a. Transformers) like ViT and Swin have emerged as a better alternative to traditional convolutional neural networks (CNNs). However, our understanding of how the new architecture…
SWAT: Scalable and Efficient Window Attention-based Transformers Acceleration on FPGAs
Efficiently supporting long context length is crucial for Transformer models. The quadratic complexity of the self-attention computation plagues traditional Transformers. Sliding window-based static sparse attention miti…
GPU