Rate this Page
★★★★★

PerTensor#

classtorch.ao.quantization.observer.PerTensor[source]#

Represents per-tensor granularity in quantization.

This granularity type calculates the quantization parametersbased off the entire tensor.