Sensitivity-Aware Finetuning for Accuracy Recovery on Deep Learning Hardware

06/05/2023
by   Lakshmi Nair, et al.
0

Existing methods to recover model accuracy on analog-digital hardware in the presence of quantization and analog noise include noise-injection training. However, it can be slow in practice, incurring high computational costs, even when starting from pretrained models. We introduce the Sensitivity-Aware Finetuning (SAFT) approach that identifies noise sensitive layers in a model, and uses the information to freeze specific layers for noise-injection training. Our results show that SAFT achieves comparable accuracy to noise-injection training and is 2x to 8x faster.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset