Build and Run QNNPACK on X86

quantization
1

Hi Pytorch team,

I want to build and run QNNPACK on X86 CPU and there are not many references to do the same. It would be great if I can get the references to build and execute QNNPACK on the X86 platform.

Thank you!

2

Look forward to your reply.

Thank you

3

Hi @Darshan , qnnpack should be included by default with x86 builds of PyTorch as long as your hardware supports it. Can you try the following in your setup:

import torch
print(torch.backends.quantized.supported_engines)
> ['qnnpack', 'none', 'onednn', 'fbgemm']