CODEWITHSUNDEEP
pythontensorflowkerasdeep-learningtensorflow2.0
Mehdi
Mehdi

Reputation: 1306

ValueError: No gradients provided for any variable: ['embedding/embeddings:0', ']

I am new in Tensorflow 2 and I want to train a multi input neural network in keras/tensorflow. This is my sample code:

First_inputs = Input(shape=(2000, ),name="first")
Second_inputs = Input(shape=(4, ),name="second")
embedding_layer = Embedding(3,3,  input_length=2000,)(First_inputs)
flatten = Flatten()(embedding_layer)
first_dense = Dense(neuronCount,kernel_initializer=initializer, )(flatten)
merge = concatenate([first_dense, Second_inputs])
drop = Dropout(dropout)(merge)
output = Dense(1, )(drop)
model = Model(inputs=[First_inputs, Second_inputs], outputs=output)
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.1,shuffle=True,  random_state=42)
First_inputs =x_train[:,0:2000]
Second_inputs =x_train[:,2000:2004]
model.fit(([First_inputs, Second_inputs], y_train),validation_data=([First_inputs, Second_inputs], y_train),verbose=1,epochs=100,steps_per_epoch=209)

However, I get this error:

ValueError: No gradients provided for any variable: ['embedding/embeddings:0', 'dense/kernel:0', 'dense/bias:0', 'dense_1/kernel:0'].

Anybody knows what the problem is? Thanks!

Upvotes: 0

Views: 514

Answers (1)

Tou You
Tou You

Reputation: 1194

Your data is numpy arrays ,you have to give two separate arguments to fit() method ,list of np.arrays as inputs and np.array as label.(remove the tuple as input):

First_inputs = Input(shape=(2000, ),name="first")
Second_inputs = Input(shape=(4, ),name="second")
embedding_layer = Embedding(3,3,  input_length=2000,)(First_inputs)
flatten = Flatten()(embedding_layer)
first_dense = Dense(neuronCount,kernel_initializer=initializer, )(flatten)
merge = concatenate([first_dense, Second_inputs])
drop = Dropout(dropout)(merge)
output = Dense(1, )(drop)
model = Model(inputs=[First_inputs, Second_inputs], outputs=output)
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.1,shuffle=True,  
random_state=42)
First_inputs =x_train[:,0:2000]
Second_inputs =x_train[:,2000:2004]
model.fit([First_inputs, Second_inputs], y_train,validation_data=([First_inputs, 
Second_inputs], y_train),verbose=1,epochs=100,steps_per_epoch=209)

Upvotes: 1

Related Questions