CODEWITHSUNDEEP
pythontensorflowmachine-learningdeep-learningkeras
hamdog
hamdog

Reputation: 1181

Keras custom loss not calculated correctly

I'm trying to use a custom loss function in Keras. My implementation looks something like:

class LossFunction:
    ...

    def loss(self, y_true, y_pred):
        ...
        localization_loss = self._localization_loss()
        confidence_loss = self._object_confidence_loss()
        category_loss = self._category_loss()

        self.loc_loss = localization_loss
        self.obj_conf_loss = confidence_loss
        self.category_loss = category_loss

        tot_loss = localization_loss + confidence_loss + category_loss
        self.tot_loss = tot_loss
        return tot_loss

I then define custom metrics to look at the stored tensors like: 

class MetricContainer:
    def __init__(self, loss_obj):
        self.loss = loss_obj

    def local_loss(self, y_true, y_pred):
        return self.loss.loc_loss

    def confidence_loss(self, y_true, y_pred):
        return self.loss.obj_conf_loss

    def category_loss(self, y_true, y_pred):
        return self.loss.category_loss

    def tot_loss(self, y_true, y_pred):
        return self.loss.tot_loss

I then compile my model with this command:

model.compile('adam', 
              loss=loss_obj.loss,
              metrics=[metric_container.local_loss, 
                       metric_container.confidence_loss, 
                       metric_container.category_loss, 
                       metric_container.tot_loss])

When I train the model (on a very small training set) I get output like:

Epoch 1/2
1/2 [==============>...............] - ETA: 76s - loss: 482.6910 - category_loss: 28.1100 - confidence_loss: 439.9192 - local_loss: 13.1180 - tot_loss: 481.1472 
2/2 [==============================] - 96s - loss: 324.6292 - category_loss: 18.1967 - confidence_loss: 296.0593 - local_loss: 8.8204 - tot_loss: 323.0764 - val_loss: 408.1170 - val_category_loss: 0.0000e+00 - val_confidence_loss: 400.0000 - val_local_loss: 6.5036 - val_tot_loss: 406.5036

For some reason tot_loss and loss don't match, even though I should be using the same value for them.

Any idea why this might be happening? Does Keras do something to modify the loss after you return it?

Upvotes: 3

Views: 786

Answers (1)

Marcin Możejko
Marcin Możejko

Reputation: 40526

Your loss is equal to a sum of a chosen loss function and regularization terms. So if you use any kind of regularization - it's affecting your loss by adding regularization terms.

Upvotes: 3

Related Questions