@@ -88,7 +88,7 @@ backPropagate :: forall shapes layers.
89which takes a network, appropriate input and target data, and returns the
89which takes a network, appropriate input and target data, and returns the
90back propagated gradients for the network. The shapes of the gradients are
90back propagated gradients for the network. The shapes of the gradients are
91−appropriate for each layer, and may be trivial for layers like `Rulu` which
91+appropriate for each layer, and may be trivial for layers like `Relu` which
92have no learnable parameters.
92have no learnable parameters.
94The gradients however can always be applied, yielding a new (hopefully better)
94The gradients however can always be applied, yielding a new (hopefully better)