✨ [Add] RepConv module in module.py

yolo/model/module.py

@@ -101,50 +101,25 @@ class SPPELAN(nn.Module):
 
 # RepVGG
 class RepConv(nn.Module):
-    # https://github.com/DingXiaoH/RepVGG
+    """A convolutional block that combines two convolution layers (kernel and point-wise)."""
+
     def __init__(
-        self, in_channels, out_channels, kernel_size=3, padding=None, stride=1, groups=1, act=nn.SiLU(), deploy=False
+        self,
+        in_channels: int,
+        out_channels: int,
+        kernel_size: _size_2_t = 3,
+        *,
+        activation: Optional[str] = "SiLU",
+        **kwargs
     ):
-
         super().__init__()
-        self.deploy = deploy
-        self.conv1 = Conv(in_channels, out_channels, kernel_size, stride=stride, groups=groups, activation=False)
-        self.conv2 = Conv(in_channels, out_channels, 1, stride=stride, groups=groups, activation=False)
-        self.act = act if isinstance(act, nn.Module) else nn.Identity()
+        self.act = get_activation(activation)
+        self.conv1 = Conv(in_channels, out_channels, kernel_size, activation=False, **kwargs)
+        self.conv2 = Conv(in_channels, out_channels, 1, activation=False, **kwargs)
 
-    def forward(self, x):
+    def forward(self, x: Tensor) -> Tensor:
         return self.act(self.conv1(x) + self.conv2(x))
 
-    def forward_fuse(self, x):
-        return self.act(self.conv(x))
-
-    # to be implement
-    # def fuse_convs(self):
-    def fuse_conv_bn(self, conv, bn):
-
-        std = (bn.running_var + bn.eps).sqrt()
-        bias = bn.bias - bn.running_mean * bn.weight / std
-
-        t = (bn.weight / std).reshape(-1, 1, 1, 1)
-        weights = conv.weight * t
-
-        bn = nn.Identity()
-        conv = nn.Conv2d(
-            in_channels=conv.in_channels,
-            out_channels=conv.out_channels,
-            kernel_size=conv.kernel_size,
-            stride=conv.stride,
-            padding=conv.padding,
-            dilation=conv.dilation,
-            groups=conv.groups,
-            bias=True,
-            padding_mode=conv.padding_mode,
-        )
-
-        conv.weight = torch.nn.Parameter(weights)
-        conv.bias = torch.nn.Parameter(bias)
-        return conv
-
 
 # ResNet
 class Res(nn.Module):