Update README.md

README.md

@@ -42,7 +42,7 @@ FAQ (as the author imagines):
 Instead of the `bottleneck` block of ResNet50 which consists of 1x1, 3x3, 1x1 in succession, this simplest version of QLNet does a 1x1, splits into two equal halves and **multiplies** them, then applies a 3x3 (depthwise), and a 1x1, *all without activation functions* except at the end of the block, where a "radial" activation function that we call `hardball` is applied.
 
 ```python
-class QLNet(nn.Module:
+class QLNet(nn.Module):
   ...
 
   def forward(self, x):
@@ -50,7 +50,7 @@ class QLNet(nn.Module:
     x = self.conv1(x)  # 1x1
     C = x.size(1) // 2
     x = x[:, :C, :, :] * x[:, C:, :, :]
-    x = self.conv2(x)  # 3x3
+    x = self.conv2(x)  # 3x3 depthwise
     x = self.conv3(x)  # 1x1
     x += x0
     if self.act3 is not None: