unable to train VGG16

[efferre99]

train_path='train'
valid_path='valid'
test_path='test'
#batches for train, valid and test set
train_batches=ImageDataGenerator(rescale =
1./255).flow_from_directory(train_path,target_size=(224,224),classes=
['dog','cat'], batch_size=10)
valid_batches=ImageDataGenerator(rescale =
1./255).flow_from_directory(valid_path,target_size=(224,224),classes=
['dog','cat'], batch_size=10)
test_batches=ImageDataGenerator(rescale =
1./255).flow_from_directory(test_path,target_size=(224,224),classes=
['dog','cat'], batch_size=10)
#build CNN
input_shape = (224, 224, 3)

model=Sequential()
model.add(Conv2D(64, (3, 3), input_shape=(224,224,3),
activation='relu',padding='same'))
model.add(Conv2D(64, (3, 3),activation='relu',padding='same'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(128, (3, 3),activation='relu',padding='same'))
model.add(Conv2D(128, (3, 3),activation='relu',padding='same'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(256, (3, 3),activation='relu',padding='same'))
model.add(Conv2D(256, (3, 3),activation='relu',padding='same'))
model.add(Conv2D(256, (3, 3),activation='relu',padding='same'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(512, (3, 3),activation='relu',padding='same'))
model.add(Conv2D(512, (3, 3),activation='relu',padding='same'))
model.add(Conv2D(512, (3, 3),activation='relu',padding='same'))
model.add(Flatten())
model.add(Dense(4096, activation='relu'))
model.add(Dense(4096, activation='relu'))

model.add(Dense(2,activation='softmax'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=
['accuracy'])

model.fit_generator(train_batches,steps_per_epoch=3,validation_data=valid_batches,validation_steps=3, epochs=5,verbose=2)
The code above is VGG16 model that classifies cats and dogs, but when I train the network I still get accuracy 0.5, why?

106s - loss: 3.9315 - acc: 0.4333 - val_loss: 8.0151 - val_acc: 0.5000
I've already used the pretrained model and it worked, but now it doesn't work

[martin-gorner]

This collection of samples does not contain a sample of VGG16. Could you please add more information about the piece of code you are commenting on ? Thanks