[Python] Why does my SimpleRNN model in Sequential API show '?' for output shape and zero...

I'm building a SimpleRNN model with an Embedding layer in Keras and encountering an issue when using the Sequential API. The model summary shows the output shape as ? and the number of trainable parameters as 0. However, when I build the same model using the Functional API, everything works as expected.

from keras.models import Sequential
from keras.layers import Embedding, SimpleRNN, Dense

model = Sequential([
Embedding(input_dim=10000, output_dim=50, input_length=100),
SimpleRNN(32),
Dense(1, activation='sigmoid')
])

model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.summary()


I get the following output:

Output shape: ?
Trainable parameters: 0(unbuilt)


However, when I use the Functional API, like this:

from keras.models import Model
from keras.layers import Input, Embedding, SimpleRNN, Dense

inputs = Input(shape=(100,))
x = Embedding(input_dim=10000, output_dim=50)(inputs)
x = SimpleRNN(32)(x)
outputs = Dense(1, activation='sigmoid')(x)

model = Model(inputs, outputs)
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.summary()


The model works fine with the correct output shape and trainable parameters.

Could anyone explain why this is happening with the Sequential API and how I can resolve this issue? Am I missing something in the configuration for Sequential?

Continue reading...