diff --git a/plinio/methods/pit/README.md b/plinio/methods/pit/README.md
index 6151135..52775cb 100644
--- a/plinio/methods/pit/README.md
+++ b/plinio/methods/pit/README.md
@@ -67,7 +67,7 @@ class Net(nn.Module):
         self.lin2 = nn.Linear()
 
 net = Net()
-exclude_types = (nn.Conv1d)
+exclude_types = (nn.Conv1d,)
 pit_net = PIT(net, exclude_types=exclude_types)
 ```
 
@@ -101,4 +101,4 @@ At the current state the optimization of the following layers is supported with
 
 When using `PIT` to optimize the receptive-field and/or the dilation of a 1D convolution, the user has two possibilities:
 - If no explicit padding is needed, use `padding='same'` for every layer with `stride=1` (PyTorch does not support same padding with stride different from 1).
-- If explicit padding is needed, use `padding='valid'` for every layer and add a `nn.ConstantPad1d` layer before each layer which need an explicit padding.
\ No newline at end of file
+- If explicit padding is needed, use `padding='valid'` for every layer and add a `nn.ConstantPad1d` layer before each layer which need an explicit padding.