Can MobileStyleGAN be deployed as openvino with fp16 accuracy?

[zhoumenghan]

Hi, thank you very much for your work. Can MobileStyleGAN be deployed as openvino with fp16 accuracy?

[bes-dev]

@zhoumenghan Thanks for your feedback. On CPU fp16 and fp32 have the same performance so it doesn't make sense to use fp16 in openvino

[zhoumenghan]

@zhoumenghan Thanks for your feedback. On CPU fp16 and fp32 have the same performance so it doesn't make sense to use fp16 in openvino

I want to use the openvino model deployed on intel GPU，so I think the FP16 model will be faster.