torch.ao.quantization.observer.default_placeholder_observer#

torch.ao.quantization.observer.default_placeholder_observer#

Default placeholder observer, usually used for quantization to torch.float16.