Dequantize stub module, before calibration, this is same as identity, this will be swapped as nnq.DeQuantize in convert.
上一页
torch.ao.quantization.stubs.QuantStub
下一页
torch.ao.quantization.stubs.QuantWrapper