Multiple parameters for 2 models - tensorflow

Question

I have 2 models which are merged into a single model. This single model is created as follows:
fm = layers.Concatenate()([m1.output, m2.output])
# create dense layer for fusion
t = layers.Dense(143, activation='relu')(fm )
d = layers.Dropout(0.5)(t)
# create softmax layer for predictions
output = layers.Dense(len(classes), activation='softmax')(d)
# create model from all those layers :)
model = keras.models.Model(inputs=[ m1.input, m2.input], outputs=output)
This single model accepts the input from the two models : these two models are
First model - 2 input parameters tx[0] and tx[1]
Second model - 1 input parameter vx
and a common Y vy
Here's how I try to use them:
history = model.fit(
[
[ np.array(tx[0]), np.array(tx[1]) ],
np.array(vx)
],
np.array(vy), verbose = 1,
validation_data = (
[
[ np.array(txv[0]), np.array(txv[1]) ],
np.array(tvv)
],
np.array(vy)),
epochs = 1200,
batch_size = 128,
callbacks = [es, mcp_save])
So my first model has 2 parameters and my second one has one parameter. They both have a common y.
But the problem is that for some reason the first parameters end up being inputted for the second model. How to resolve that?

Answer 1

If I understand your model correctly, it has three inputs layers and you only mentioned two in your final model.
Use the variable name of each of the two input_layers of the first model instead of m1.input.
#the two input layers of the first Model m1:
inputA = tf.keras.Input(shape=(32,))
inputB = tf.keras.Input(shape=(128,))
#the input layer of the second Model m2:
inputC = tf.keras.Input(shape=(128,))
#after merging the two models m1 and m2 into one Model model :
model = keras.models.Model(inputs=[inputA, inputB , inputC], outputs=output)