What is the difference between the predictive and predicted methods of the Keras model?

Question

According to keras documentation :

predict_on_batch(self, x) Returns predictions for a single batch of samples. 

However, when invoking a package, there is no difference with the standard predict method, whether with one or more elements.

 model.predict_on_batch(np.zeros((n, d_in))) 

coincides with

 model.predict(np.zeros((n, d_in))) 

(a numpy.ndarray form (n, d_out )

Answer 1

The difference is that you transmit as x data exceeding one batch.

predict will go through all the data, batch, predictive labels. Thus, he internally performs batch separation and feeds one batch at a time.

predict_on_batch , on the other hand, assumes that the data you transfer is exactly one batch and, thus, transfers it to the network. It will not try to break it (which, depending on your installation, may be problematic for your GPU memory if the array is very large)

Answer 2

I just want to add something that doesn't match the comment. Predict seems to check carefully output form:

 class ExtractShape(keras.engine.topology.Layer): def call(self, x): return keras.backend.sum(x, axis=0) def compute_output_shape(self, input_shape): return input_shape a = keras.layers.Input((None, None)) b = ExtractShape()(a) m = keras.Model(a, b) m.compile(optimizer=keras.optimizers.Adam(), loss='binary_crossentropy') A = np.ones((5,4,3)) 

Then:

 In [163]: m.predict_on_batch(A) Out[163]: array([[5., 5., 5.], [5., 5., 5.], [5., 5., 5.], [5., 5., 5.]], dtype=float32) In [164]: m.predict_on_batch(A).shape Out[164]: (4, 3) 

But:

 In [165]: m.predict(A) --------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-165-c5ba5fc88b6e> in <module>() ----> 1 m.predict(A) ~/miniconda3/envs/ccia/lib/python3.6/site-packages/keras/engine/training.py in predict(self, x, batch_size, verbose, steps) 1746 f = self.predict_function 1747 return self._predict_loop(f, ins, batch_size=batch_size, -> 1748 verbose=verbose, steps=steps) 1749 1750 def train_on_batch(self, x, y, ~/miniconda3/envs/ccia/lib/python3.6/site-packages/keras/engine/training.py in _predict_loop(self, f, ins, batch_size, verbose, steps) 1306 outs.append(np.zeros(shape, dtype=batch_out.dtype)) 1307 for i, batch_out in enumerate(batch_outs): -> 1308 outs[i][batch_start:batch_end] = batch_out 1309 if verbose == 1: 1310 progbar.update(batch_end) ValueError: could not broadcast input array from shape (4,3) into shape (5,3) 

I am not sure if this is really a mistake.