Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference

• URL: http://arxiv.org/abs/2306.14393v1
• Date: Mon, 26 Jun 2023 03:06:57 GMT
• Title: Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference
• Authors: Junyan Li, Li Lyna Zhang, Jiahang Xu, Yujing Wang, Shaoguang Yan, Yunqing Xia, Yuqing Yang, Ting Cao, Hao Sun, Weiwei Deng, Qi Zhang, Mao Yang
• Abstract summary: ToP is a ranking-distilled token distillation technique, which distills effective token rankings from the final layer of unpruned models to early layers of pruned models. ToP reduces the average FLOPs of BERT by 8.1x while achieving competitive accuracy on GLUE, and provides a real latency speedup of up to 7.4x on an Intel CPU.
• Score: 18.308180927492643
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Deploying pre-trained transformer models like BERT on downstream tasks in resource-constrained scenarios is challenging due to their high inference cost, which grows rapidly with input sequence length. In this work, we propose a constraint-aware and ranking-distilled token pruning method ToP, which selectively removes unnecessary tokens as input sequence passes through layers, allowing the model to improve online inference speed while preserving accuracy. ToP overcomes the limitation of inaccurate token importance ranking in the conventional self-attention mechanism through a ranking-distilled token distillation technique, which distills effective token rankings from the final layer of unpruned models to early layers of pruned models. Then, ToP introduces a coarse-to-fine pruning approach that automatically selects the optimal subset of transformer layers and optimizes token pruning decisions within these layers through improved $L_0$ regularization. Extensive experiments on GLUE benchmark and SQuAD tasks demonstrate that ToP outperforms state-of-the-art token pruning and model compression methods with improved accuracy and speedups. ToP reduces the average FLOPs of BERT by 8.1x while achieving competitive accuracy on GLUE, and provides a real latency speedup of up to 7.4x on an Intel CPU.

Related papers

• FiRST: Finetuning Router-Selective Transformers for Input-Adaptive Latency Reduction [11.146015814220858]
  FIRST is an algorithm that reduces inference latency by using layer-specific routers to select a subset of transformer layers adaptively for each input sequence. Our approach reveals that input adaptivity is critical - indeed, different task-specific middle layers play a crucial role in evolving hidden representations depending on task.
  arXiv  Detail & Related papers  (2024-10-16T12:45:35Z)
• DopQ-ViT: Towards Distribution-Friendly and Outlier-Aware Post-Training Quantization for Vision Transformers [2.0862654518798034]
  We propose a Distribution-Friendly and Outlier-Aware Post-training Quantization method for Vision Transformers. DopQ-ViT analyzes the inefficiencies of current quantizers and introduces a distribution-friendly Tan Quantizer called TanQ. DopQ-ViT has been extensively validated and significantly improves the performance of quantization models.
  arXiv  Detail & Related papers  (2024-08-06T16:40:04Z)
• Sparse is Enough in Fine-tuning Pre-trained Large Language Models [98.46493578509039]
  We propose a gradient-based sparse fine-tuning algorithm, named Sparse Increment Fine-Tuning (SIFT) We validate its effectiveness on a range of tasks including the GLUE Benchmark and Instruction-tuning.
  arXiv  Detail & Related papers  (2023-12-19T06:06:30Z)
• DPBERT: Efficient Inference for BERT based on Dynamic Planning [11.680840266488884]
  Existing input-adaptive inference methods fail to take full advantage of the structure of BERT. We propose Dynamic Planning in BERT, a novel fine-tuning strategy that can accelerate the inference process of BERT. Our method reduces latency to 75% while maintaining 98% accuracy, yielding a better accuracy-speed trade-off compared to state-of-the-art input-adaptive methods.
  arXiv  Detail & Related papers  (2023-07-26T07:18:50Z)
• Approximated Prompt Tuning for Vision-Language Pre-trained Models [54.326232586461614]
  In vision-language pre-trained models, prompt tuning often requires a large number of learnable tokens to bridge the gap between the pre-training and downstream tasks. We propose a novel Approximated Prompt Tuning (APT) approach towards efficient VL transfer learning.
  arXiv  Detail & Related papers  (2023-06-27T05:43:47Z)
• Pruning Deep Neural Networks from a Sparsity Perspective [34.22967841734504]
  Pruning is often achieved by dropping redundant weights, neurons, or layers of a deep network while attempting to retain a comparable test performance. We propose PQ Index (PQI) to measure the potential compressibility of deep neural networks and use this to develop a Sparsity-informed Adaptive Pruning (SAP) algorithm.
  arXiv  Detail & Related papers  (2023-02-11T04:52:20Z)
• Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT [22.904252855587348]
  We propose a fine- and coarse-granularity hybrid self-attention that reduces the cost through progressively shortening the computational sequence length in self-attention. We show that FCA offers a significantly better trade-off between accuracy and FLOPs compared to prior methods.
  arXiv  Detail & Related papers  (2022-03-17T03:33:47Z)
• Layer Pruning on Demand with Intermediate CTC [50.509073206630994]
  We present a training and pruning method for ASR based on the connectionist temporal classification (CTC) We show that a Transformer-CTC model can be pruned in various depth on demand, improving real-time factor from 0.005 to 0.002 on GPU.
  arXiv  Detail & Related papers  (2021-06-17T02:40:18Z)
• Accelerating BERT Inference for Sequence Labeling via Early-Exit [65.7292767360083]
  We extend the recent successful early-exit mechanism to accelerate the inference of PTMs for sequence labeling tasks. We also propose a token-level early-exit mechanism that allows partial tokens to exit early at different layers. Our approach can save up to 66%-75% inference cost with minimal performance degradation.
  arXiv  Detail & Related papers  (2021-05-28T14:39:26Z)
• FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization [72.9385528828306]
  A typical transducer model decodes the output sequence conditioned on the current acoustic state. The number of blank tokens in the prediction results accounts for nearly 90% of all tokens. We propose a method named fast-skip regularization, which tries to align the blank position predicted by a transducer with that predicted by a CTC model.
  arXiv  Detail & Related papers  (2021-04-07T03:15:10Z)
• BERT Loses Patience: Fast and Robust Inference with Early Exit [91.26199404912019]
  We propose Patience-based Early Exit as a plug-and-play technique to improve the efficiency and robustness of a pretrained language model. Our approach improves inference efficiency as it allows the model to make a prediction with fewer layers.
  arXiv  Detail & Related papers  (2020-06-07T13:38:32Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.