I have a model defined as such:
model = Model(inputs=[inputs, y_true, is_weight], outputs=[y_pred])
I want to specify the training and validation data, since I am performing k-fold cross-validation and I have predefined the splits beforehand. I fit the model as follows:
model.fit([X_train, Y_train, wgt_map_train], batch_size=1, epochs=200,
validation_data = [X_val, Y_val, wgt_map_val],
callbacks = [checkpointer, earlystopper])
The shapes of the input both for training and validation are the same, except for the first dimension:
X_train.shape = (6732, 12, 86, 98, 1)
X_val.shape = (765, 12, 86, 98, 1)
etc.
However, when I run the code, I get the following error:
~/miniconda3/envs/segment/lib/python3.7/site-packages/keras/engine/training.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)
970 val_x, val_y,
971 sample_weight=val_sample_weight,
--> 972 batch_size=batch_size)
973 if self._uses_dynamic_learning_phase():
974 val_ins = val_x + val_y + val_sample_weights + [0.]
~/miniconda3/envs/segment/lib/python3.7/site-packages/keras/engine/training.py in _standardize_user_data(self, x, y, sample_weight, class_weight, check_array_lengths, batch_size)
749 feed_input_shapes,
750 check_batch_axis=False, # Don't enforce the batch size.
--> 751 exception_prefix='input')
752
753 if y is not None:
~/miniconda3/envs/segment/lib/python3.7/site-packages/keras/engine/training_utils.py in standardize_input_data(data, names, shapes, check_batch_axis, exception_prefix)
100 'Expected to see ' + str(len(names)) + ' array(s), '
101 'but instead got the following list of ' +
--> 102 str(len(data)) + ' arrays: ' + str(data)[:200] + '...')
103 elif len(names) > 1:
104 raise ValueError(
ValueError: Error when checking model input: the list of Numpy arrays that you are passing to your model is not the size the model expected. Expected to see 3 array(s), but instead got the following list of 1 arrays: [array([[[[[0. ],
[0. ],
[0. ],
...,
[0. ],
[0. ],
[0. ]],
[[0. ],
[0. ...
If I remove validation_data and I replace it for validation_split = 0.1 works, but then this would not be an adequate way of doing k-fold CV. Thanks!
EDIT I have also tried adding 'validation_batch_size = 1' in model.fit, but then I get prompted with:
TypeError Traceback (most recent call last)
<ipython-input-46-eefd2297c992> in <module>
42 validation_data = [X_val, Y_val, wgt_map_val],
43 validation_batch_size = 1,
---> 44 callbacks = [checkpointer, earlystopper,tb])
45 time2=time()
46
~/miniconda3/envs/segment/lib/python3.7/site-packages/keras/engine/training.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)
940 epochs = kwargs.pop('nb_epoch')
941 if kwargs:
--> 942 raise TypeError('Unrecognized keyword arguments: ' + str(kwargs))
943 if x is None and y is None and steps_per_epoch is None:
944 raise ValueError('If fitting from data tensors, '
TypeError: Unrecognized keyword arguments: {'validation_batch_size': 1}
I have checked similar posts like: Deep Learning fit error (the list of Numpy arrays that you are passing to your model is not the size the model expected.) and Keras functional model with multiple inputs
