@@ -177,7 +177,9 @@ We also need that the last activation layer of the network to be a softmax layer
-`labels_train` a vector of size `batch_size`, and
-`learning_rate` the learning rate,
that perform one gradient descent step using a binary cross-entropy loss. We admit that $`\frac{\partial C}{\partial Z^{(2)}} = A^{(2)} - Y`$, where $`Y`$ is a one-hot vector encoding the label. The function must return:
that perform one gradient descent step using a binary cross-entropy loss.
We admit that $`\frac{\partial C}{\partial Z^{(2)}} = A^{(2)} - Y`$, where $`Y`$ is a one-hot vector encoding the label.
The function must return:
- `w1`, `b1`, `w2` and `b2` the updated weights and biases of the network,
- `loss` the loss, for monitoring purpose.
13. Write the function `train_mlp` taking as parameters:
