用卷积网络加速长对话理解，速度比Transformer快80%以上
An Empirical Evaluation of Encoder Architectures for Fast Real-Time Long Conversational Understanding
- 采用卷积神经网络替代传统Transformer处理长对话
- 训练快2.6倍，推理快80%，内存节省72%
- 适合需要实时处理长文本的客服系统场景
分析客户通话记录等长文本数据成本高且耗时。现有基于Transformer的机器学习方法虽能建模交互，但受限于固定长度输入和二次方复杂度的自注意力机制，难以应用于长序列任务，尤其在实时场景中。本文评估了近期提出的高效Transformer变体（如Performer、Reformer）及一种基于CNN的架构，在实时与近实时长对话理解任务中的表现。结果表明，卷积模型平均训练速度快2.6倍，推理速度快80%，内存效率提升72%。此外，通过Long Range Arena基准测试，该模型在通用长文档分析任务中表现出竞争力。
原文摘要 · Abstract (English)
Analyzing long text data such as customer call transcripts is a cost-intensive and tedious task. Machine learning methods, namely Transformers, are leveraged to model agent-customer interactions. Unfortunately, Transformers adhere to fixed-length architectures and their self-attention mechanism scales quadratically with input length. Such limitations make it challenging to leverage traditional Transformers for long sequence tasks, such as conversational understanding, especially in real-time use cases. In this paper we explore and evaluate recently proposed efficient Transformer variants (e.g. Performer, Reformer) and a CNN-based architecture for real-time and near real-time long conversational understanding tasks. We show that CNN-based models are dynamic, ~2.6x faster to train, ~80% faster inference and ~72% more memory efficient compared to Transformers on average. Additionally, we evaluate the CNN model using the Long Range Arena benchmark to demonstrate competitiveness in general long document analysis.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。