Rate this Page

ConvAddReLU2d#

class torch.ao.nn.intrinsic.modules.fused.ConvAddReLU2d(conv, add, relu)[source]#

This is a sequential container which calls the Conv2d, add, Relu. During quantization this will be replaced with the corresponding fused module.

forward(x1, x2)[source]#

Applies convolution to x1, adds the result to x2, and applies ReLU.