Long Range Arena: A Benchmark for Efficient Transformers

Abstract

Transformers do not scale very well to long sequence lengths largely because of quadratic self-attention complexity. In the recent months, a wide spectrum of efficient, fast Transformers have been proposed to tackle this problem, more often than not claiming superior or comparable model quality to vanilla Transformer models. To this date, there is no well-established consensus on how to evaluate this class of models. Moreover, inconsistent benchmarking on a wide spectrum of tasks and datasets makes it difficult to assess relative model quality amongst many models. This paper proposes a systematic and unified benchmark, LRA, specifically focused on evaluating model quality under long-context scenarios. Our benchmark is a suite of tasks consisting of sequences ranging from $1K$ to $16K$ tokens, encompassing a wide range of data types and modalities such as text, natural, synthetic images, and mathematical expressions requiring similarity, structural, and visual-spatial reasoning. We systematically evaluate ten well-established long-range Transformer models (Reformers, Linformers, Linear Transformers, Sinkhorn Transformers, Performers, Synthesizers, Sparse Transformers, and Longformers) on our newly proposed benchmark suite. LRA paves the way towards better understanding this class of efficient Transformer models, facilitates more research in this direction, and presents new challenging tasks to tackle. Our benchmark code will be released at https://github.com/google-research/long-range-arena.

References in corpus (16)

• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• Longformer: The Long-Document Transformer
• Linformer: Self-Attention with Linear Complexity
• A Deep Relevance Matching Model for Ad-hoc Retrieval
• REALM: Retrieval-Augmented Language Model Pre-Training
• Generating Long Sequences with Sparse Transformers
• Axial Attention in Multidimensional Transformers
• Reformer: The Efficient Transformer
• Big Bird: Transformers for Longer Sequences
• Synthesizer: Rethinking Self-Attention in Transformer Models
• Rethinking Attention with Performers
• Beyond 512 Tokens: Siamese Multi-depth Transformer-based Hierarchical Encoder for Long-Form Document Matching
• Sparse Sinkhorn Attention
• Generating Wikipedia by Summarizing Long Sequences
• Masked Language Modeling for Proteins via Linearly Scalable Long-Context Transformers
• ETC: Encoding Long and Structured Inputs in Transformers

Cited by in corpus (30)

• An Empirical Survey on Long Document Summarization: Datasets, Models and Metrics
• Pretrained Transformers as Universal Computation Engines
• Vector database management systems: Fundamental concepts, use-cases, and current challenges
• Chasing Sparsity in Vision Transformers: An End-to-End Exploration
• Long-Range Transformers for Dynamic Spatiotemporal Forecasting
• TeraPipe: Token-Level Pipeline Parallelism for Training Large-Scale Language Models
• Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention
• Text Guide: Improving the quality of long text classification by a text selection method based on feature importance
• Efficient Quantized Sparse Matrix Operations on Tensor Cores
• Match-Ignition: Plugging PageRank into Transformer for Long-form Text Matching
• The NLP Task Effectiveness of Long-Range Transformers
• Fourier Transformer: Fast Long Range Modeling by Removing Sequence Redundancy with FFT Operator
• Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search
• Scatterbrain: Unifying Sparse and Low-rank Attention Approximation
• Co-BERT: A Context-Aware BERT Retrieval Model Incorporating Local and Query-specific Context
• The Benchmark Lottery
• Pixelated Butterfly: Simple and Efficient Sparse training for Neural Network Models
• Eigen Analysis of Self-Attention and its Reconstruction from Partial Computation
• Tracking Without Re-recognition in Humans and Machines
• The Role of Global and Local Context in Named Entity Recognition
• Sparse Factorization of Large Square Matrices
• Linear Self-Attention Approximation via Trainable Feedforward Kernel
• Updater-Extractor Architecture for Inductive World State Representations
• The Efficiency Misnomer
• Knowledge Transfer by Discriminative Pre-training for Academic Performance Prediction
• SHORING: Design Provable Conditional High-Order Interaction Network via Symbolic Testing
• Metadata Shaping: Natural Language Annotations for the Tail
• Diagnosing and exploiting the computational demands of videos games for deep reinforcement learning
• Semantic maps and metrics for science Semantic maps and metrics for science using deep transformer encoders
• Ripple Attention for Visual Perception with Sub-quadratic Complexity