ConvBn2d¶
- class ConvBn2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, conv_mode='cross_correlation', compute_mode='default', eps=1e-05, momentum=0.9, affine=True, track_running_stats=True, padding_mode='zeros', **kwargs)[源代码]¶
融合了
Conv2d和BatchNorm2d的Module. 可以调用quantize_qat替换成QATModule型的ConvBn2d.