Skip to main content
. 2022 Jun 1;2022:8039281. doi: 10.1155/2022/8039281

Table 11.

Accuracy comparison after quantization.

ImageNet100 Original model (acc.) Pruned model (acc.)
FP32 INT8 FP32 INT8
VGG13BN 78.85 78.05 77.85 77.23
ResNet101 79.08 77.74 78.28 76.44