Do I need to specify an activation function after a Conv1d?

I assume that I do, but I wanted to make sure since I don’t see any activations in the docs?

There’s no nonlinearity built in to nn.Conv1d. You’ll need to add an activation function to your network if you need a nonlinearity.