Keras: TypeError with batch_size

Question

I'm using Keras (with Python 3.6) to predict the output of an array (x_test), but I get a TypeError in return.

Here's my code for the prediction:

x_test = [[8],[6],[0],[2],[0],[0],[0],[0],[112.128],[0],[0],[2],[0],[1],[1],[2],[2]]
prediction = model.predict(model, x_test, batch_size = 32, verbose = 1)

And here's the error I get:

TypeError                                 Traceback (most recent call last)
<ipython-input-14-286495dc15a7> in <module>()
  1 x_test = [[8],[6],[0],[2],[0],[0],[0],[0],[112.128],[0],[0],[2],[0],[1],[1],[2],[2]]
  2 
----> 3 prediction = model.predict(model, x_test, batch_size =(17,1), verbose = 1)

TypeError: predict() got multiple values for argument 'batch_size'

If anybody has any advice on what went wrong, any help is greatly appreciated.

For reference, here's my neural network, which seems to work fine.

model = Sequential()

model.add(Dense(32, input_dim=17, init='uniform', activation='relu' ))
model.add(Dense(64, init='uniform', activation='relu'))
model.add(Dense(128, init='uniform', activation='relu'))
model.add(Dense(64, init='uniform', activation='sigmoid'))
model.add(Dense(32, init='uniform', activation='sigmoid'))
model.add(Dense(16, init='uniform', activation='sigmoid'))
model.add(Dense(8, init='uniform', activation='sigmoid'))
model.add(Dense(4, init='uniform', activation='sigmoid'))
model.add(Dense(1, init='uniform', activation='sigmoid'))

# Compile model
model.compile(loss='mean_squared_logarithmic_error', optimizer='SGD', metrics=['accuracy'])

# Fit model
history = model.fit(X, Y, nb_epoch=300, validation_split=0.2, batch_size=3)

Many thanks!

Answer 1

You don't need to pass the model argument in model.predict, since the default for predict is predict(self, x, batch_size=32, verbose=0) which model is automatically defined by self.

So your code should just be like:

prediction = model.predict(x_test, batch_size = 32, verbose = 1)

And according to the documentation, x should be a numpy.array not a list.

Arguments:

x: the input data, as a Numpy array.

batch_size: integer.

verbose: verbosity mode, 0 or 1.

Which means that x_test should instead be:

x_test = np.array([[8],[6],[0],[2],[0],[0],[0],[0],[112.128],[0],[0],[2],[0],[1],[1],[2],[2]])