Rate this Page
★★★★★

get_block_size#

classtorch.ao.quantization.observer.get_block_size(input_shape,granularity)[source]#

Get the block size based on the input shape and granularity type.

Parameters

  • input_shape (tuple[int,...]) – The input tensor shape possibly more than 2 dimensions

  • granularity (Granularity) – The granularity type of the quantization

Return type

tuple[int, …]