Does Keras ignore labels of masked values?

Question

I'm implementing a LSTM model with Keras. I padded my sequences to a certain length to feed the dataset in the right way into the model.

At the moment, my model is the following:

model = tf.keras.Sequential()
model.add(Masking(mask_value=0., input_shape=(timesteps, features)))
model.add(LSTM(units=100, return_sequences=True, input_shape=(timesteps, features)))
model.add(Dropout(0.5))
model.add(Dense(1, activation='sigmoid'))

Does Keras automatically skip labels of masked values in the loss function?

Answer 1

Yes, if your model utilizes masking then the objective function (i.e. loss function) would be automatically augmented to support masking and therefore ignoring masked samples/timesteps in calculation of loss. Actually, weighted_masked_objective is the function which does this under the hood:

def weighted_masked_objective(fn):
    """Adds support for masking and sample-weighting to an objective function.
    It transforms an objective function `fn(y_true, y_pred)`
    into a sample-weighted, cost-masked objective function
    `fn(y_true, y_pred, weights, mask)`.
    # Arguments
        fn: The objective function to wrap,
            with signature `fn(y_true, y_pred)`.
    # Returns
        A function with signature `fn(y_true, y_pred, weights, mask)`.
    """
    if fn is None:
        return None

    def weighted(y_true, y_pred, weights, mask=None):
        """Wrapper function.
        # Arguments
            y_true: `y_true` argument of `fn`.
            y_pred: `y_pred` argument of `fn`.
            weights: Weights tensor.
            mask: Mask tensor.
        # Returns
            Scalar tensor.
        """
        # score_array has ndim >= 2
        score_array = fn(y_true, y_pred)
        if mask is not None:
            # Cast the mask to floatX to avoid float64 upcasting in Theano
            mask = K.cast(mask, K.floatx())
            # mask should have the same shape as score_array
            score_array *= mask
            #  the loss per batch should be proportional
            #  to the number of unmasked samples.
            score_array /= K.mean(mask) + K.epsilon()

        # apply sample weighting
        if weights is not None:
            # reduce score_array to same ndim as weight array
            ndim = K.ndim(score_array)
            weight_ndim = K.ndim(weights)
            score_array = K.mean(score_array,
                                 axis=list(range(weight_ndim, ndim)))
            score_array *= weights
            score_array /= K.mean(K.cast(K.not_equal(weights, 0), K.floatx()))
        return K.mean(score_array)
    return weighted