DeepAI AI Chat
Log In Sign Up

Pruning Redundant Mappings in Transformer Models via Spectral-Normalized Identity Prior

by   Zi Lin, et al.

Traditional (unstructured) pruning methods for a Transformer model focus on regularizing the individual weights by penalizing them toward zero. In this work, we explore spectral-normalized identity priors (SNIP), a structured pruning approach that penalizes an entire residual module in a Transformer model toward an identity mapping. Our method identifies and discards unimportant non-linear mappings in the residual connections by applying a thresholding operator on the function norm. It is applicable to any structured module, including a single attention head, an entire attention block, or a feed-forward subnetwork. Furthermore, we introduce spectral normalization to stabilize the distribution of the post-activation values of the Transformer layers, further improving the pruning effectiveness of the proposed methodology. We conduct experiments with BERT on 5 GLUE benchmark tasks to demonstrate that SNIP achieves effective pruning results while maintaining comparable performance. Specifically, we improve the performance over the state-of-the-art by 0.5 to 1.0


Layer-wise Pruning of Transformer Attention Heads for Efficient Language Modeling

While Transformer-based models have shown impressive language modeling p...

Pruning a BERT-based Question Answering Model

We investigate compressing a BERT-based question answering system by pru...

MLPruning: A Multilevel Structured Pruning Framework for Transformer-based Models

Pruning is an effective method to reduce the memory footprint and comput...

Differentiable Subset Pruning of Transformer Heads

Multi-head attention, a collection of several attention mechanisms that ...

Entangled Residual Mappings

Residual mappings have been shown to perform representation learning in ...

Pruning Attention Heads of Transformer Models Using A* Search: A Novel Approach to Compress Big NLP Architectures

Recent years have seen a growing adoption of Transformer models such as ...