AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning
University of Wisconsin-Madison
arXiv:2102.01386 [cs.LG], (2 Feb 2021)
@misc{liu2021autofreeze,
title={AutoFreeze: Automatically Freezing Model Blocks to Accelerate Fine-tuning},
author={Yuhan Liu and Saurabh Agarwal and Shivaram Venkataraman},
year={2021},
eprint={2102.01386},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
With the rapid adoption of machine learning (ML), a number of domains now use the approach of fine-tuning models pre-trained on a large corpus of data. However, our experiments show that even fine-tuning on models like BERT can take many hours when using GPUs. While prior work proposes limiting the number of layers that are fine-tuned, e.g., freezing all layers but the last layer, we find that such static approaches lead to reduced accuracy. We propose, AutoFreeze, a system that uses an adaptive approach to choose which layers are trained and show how this can accelerate model fine-tuning while preserving accuracy. We also develop mechanisms to enable efficient caching of intermediate activations which can reduce the forward computation time when performing fine-tuning. Our evaluation on fourNLP tasks shows that AutoFreeze, with caching enabled, can improve fine-tuning performance by up to 2.55x.
February 7, 2021 by hgpu
