SEVEN: Pruning Transformer Model by Reserving Sentinels

Abstract

Large-scale Transformer models (TM) have demonstrated outstanding performance across various tasks. However, their considerable parameter size restricts their applicability, particularly on mobile devices. Due to the dynamic and intricate nature of gradients on TM compared to Convolutional Neural Networks, commonly used pruning methods tend to retain weights with larger gradient noise. This results in pruned models that are sensitive to sparsity and datasets, exhibiting suboptimal performance. Symbolic Descent (SD) is a general approach for training and fine-tuning TM. In this paper, we attempt to describe the noisy batch gradient sequences on TM through the cumulative process of SD. We utilize this design to dynamically assess the importance scores of weights.SEVEN is introduced by us, which particularly favors weights with consistently high sensitivity, i.e., weights with small gradient noise. These weights are tended to be preserved by SEVEN. Extensive experiments on various TM in natural language, question-answering, and image classification domains are conducted to validate the effectiveness of SEVEN. The results demonstrate significant improvements of SEVEN in multiple pruning scenarios and across different sparsity levels. Additionally, SEVEN exhibits robust performance under various fine-tuning strategies. The code is publicly available at https://github.com/xiaojinying/SEVEN.

IJCNN 2024

References in corpus (11)

• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
• The State of Sparsity in Deep Neural Networks
• Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model
• Symbolic Discovery of Optimization Algorithms
• Structured Pruning of Large Language Models
• Progressive Skeletonization: Trimming more fat from a network at initialization
• Why are Adaptive Methods Good for Attention Models?
• Prune Once for All: Sparse Pre-Trained Language Models
• The Topological BERT: Transforming Attention into Topology for Natural Language Processing
• LNPT: Label-free Network Pruning and Training

Cited by in corpus (1)

• LNPT: Label-free Network Pruning and Training