ValueError: Input 0 of layer "sequential_35" is incompatible with the layer: expected shape=(None, 800, 1, 100), found shape=(None, 1, 100)

Question

Hi Im currently working on the below program but im getting below error: ValueError: Input 0 of layer "sequential_35" is incompatible with the layer: expected shape=(None, 800, 1, 100), found shape=(None, 1, 100). I need to convert it to 4D output to run the program. Please help. Advance Thank You.

      
tok = Tokenizer(num_words=max_words)
tok.fit_on_texts(X)
print('Found %s unique tokens.' % len(tok.word_index))
X = tok.texts_to_sequences(X.values)
X = sequence.pad_sequences(X, maxlen=max_len)
print('Shape of data tensor:', X.shape)

X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.15)

le = LabelEncoder()
Y_train_enc = le.fit_transform(Y_train)
Y_train_enc = np_utils.to_categorical(Y_train_enc)

Y_test_enc = le.transform(Y_test)
Y_test_enc = np_utils.to_categorical(Y_test_enc)


def malware_model(act_func="softsign"):
    model = Sequential()
    model.add(Embedding(257, 128, input_shape=(max_words, 1, max_len)))
    model.add(ConvLSTM2D(filters=12, kernel_size=(1, 2),
              dropout=0.1, activation='relu', return_sequences=True))
    model.add(ConvLSTM2D(filters=12, kernel_size=(
        1, 2), dropout=0.1, activation='relu'))
    model.add(ConvLSTM1D(filters=12, kernel_size=(
        1), dropout=0.1, activation='relu'))
    model.add(Dense(128, activation=act_func))
    model.add(Dropout(0.1))
    model.add(Dense(256, activation=act_func))
    model.add(Dropout(0.1))
    model.add(Dense(128, activation=act_func))
    model.add(Dropout(0.1))
    model.add(Dense(1, name='out_layer', activation="linear"))
    return model


model = malware_model()
print(model.summary())
model.compile(loss='mse', optimizer="rmsprop",
              metrics=['accuracy'])


history = model.fit(tf.expand_dims(X_train, axis=1), batch_size=1000, epochs=10,
                    validation_data=(X_test, Y_test), verbose=1)`

    ValueError: Input 0 of layer "sequential_35" is incompatible with the layer: expected shape=(None, 800, 1, 100), found shape=(None, 1, 100)'

    

Answer 1

Looks like issue with train data shape. Make sure it should match with input shape of first layer.

Below code snippet works

import tensorflow as tf 
max_words= 800
max_len =100

model = tf.keras.Sequential()
model.add(tf.keras.layers.Embedding(257, 128, input_shape=(max_words, 1, max_len)))
model.add(tf.keras.layers.ConvLSTM2D(filters=12, kernel_size=(1, 2),
              dropout=0.1, activation='relu', return_sequences=True))
model.add(tf.keras.layers.ConvLSTM2D(filters=12, kernel_size=(
        1, 2), dropout=0.1, activation='relu'))
model.add(tf.keras.layers.ConvLSTM1D(filters=12, kernel_size=(
        1), dropout=0.1, activation='relu'))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.1))
model.add(tf.keras.layers.Dense(256, activation='relu'))
model.add(tf.keras.layers.Dropout(0.1))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.1))
model.add(tf.keras.layers.Dense(1, name='out_layer', activation="linear"))

model.summary()

Output

Model: "sequential_2"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 embedding_1 (Embedding)     (None, 800, 1, 100, 128)  32896     
                                                                 
 conv_lstm2d_2 (ConvLSTM2D)  (None, 800, 1, 99, 12)    13488     
                                                                 
 conv_lstm2d_3 (ConvLSTM2D)  (None, 1, 98, 12)         2352      
                                                                 
 conv_lstm1d_1 (ConvLSTM1D)  (None, 98, 12)            1200      
                                                                 
 dense (Dense)               (None, 98, 128)           1664      
                                                                 
 dropout (Dropout)           (None, 98, 128)           0         
                                                                 
 dense_1 (Dense)             (None, 98, 256)           33024     
                                                                 
 dropout_1 (Dropout)         (None, 98, 256)           0         
                                                                 
 dense_2 (Dense)             (None, 98, 128)           32896     
                                                                 
 dropout_2 (Dropout)         (None, 98, 128)           0         
                                                                 
 out_layer (Dense)           (None, 98, 1)             129       
                                                                 
=================================================================
Total params: 117,649
Trainable params: 117,649
Non-trainable params: 0
_______________________