Rate this Page

get_default_qconfig_mapping#

class torch.ao.quantization.qconfig_mapping.get_default_qconfig_mapping(backend='x86', version=0)[source]#

Return the default QConfigMapping for post training quantization.

Parameters
  • backend (*) – the quantization backend for the default qconfig mapping, should be one of [“x86” (default), “fbgemm”, “qnnpack”, “onednn”]

  • version (*) – the version for the default qconfig mapping

Return type

QConfigMapping