Fourier Transformer: Fast Long Range Modeling by Removing Sequence Redundancy with FFT Operator

Abstract

The transformer model is known to be computationally demanding, and prohibitively costly for long sequences, as the self-attention module uses a quadratic time and space complexity with respect to sequence length. Many researchers have focused on designing new forms of self-attention or introducing new parameters to overcome this limitation, however a large portion of them prohibits the model to inherit weights from large pretrained models. In this work, the transformer's inefficiency has been taken care of from another perspective. We propose Fourier Transformer, a simple yet effective approach by progressively removing redundancies in hidden sequence using the ready-made Fast Fourier Transform (FFT) operator to perform Discrete Cosine Transformation (DCT). Fourier Transformer is able to significantly reduce computational costs while retain the ability to inherit from various large pretrained models. Experiments show that our model achieves state-of-the-art performances among all transformer-based models on the long-range modeling benchmark LRA with significant improvement in both speed and space. For generative seq-to-seq tasks including CNN/DailyMail and ELI5, by inheriting the BART weights our model outperforms the standard BART and other efficient models. Our code is publicly available at https://github.com/LUMIA-Group/FourierTransformer

References in corpus (15)

• Longformer: The Long-Document Transformer
• A Structured Self-attentive Sentence Embedding
• Linformer: Self-Attention with Linear Complexity
• The Pile: An 800GB Dataset of Diverse Text for Language Modeling
• Axial Attention in Multidimensional Transformers
• Reducing Transformer Depth on Demand with Structured Dropout
• Efficient Transformers: A Survey
• Synthesizer: Rethinking Self-Attention in Transformer Models
• Long Range Arena: A Benchmark for Efficient Transformers
• Random Feature Attention
• Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing
• Charformer: Fast Character Transformers via Gradient-based Subword Tokenization
• Long-Short Transformer: Efficient Transformers for Language and Vision
• Luna: Linear Unified Nested Attention
• Knowledge Infused Decoding