Failed to find data adapter that can handle input:...(list to numpy array with given dimensions)

Question

I have a list X in python which contains 20000 images of dimension (100,100,3) each. Each of the images is of type numpy.ndarray. I wish to input this list of images as the training vector for a CNN model using model.fit(X) in Tensorflow. However doing so gives me the following error:

ValueError: Failed to find data adapter that can handle input: (<class 'list'> containing values of types {"<class 'numpy.ndarray'>"}), (<class 'list'> containing values of types {"<class 'numpy.int64'>"})

Would converting list X into a numpy.ndarray of dimensions (20000,100,100,3) help?

If yes, how can I do that?

If not, could you please suggest what can be done?

What does np.array(X) produce? Especially what's the shape and dtype? Beware of 1d array with object dtype - the means the list contains arrays that differ in shape. — hpaulj
– hpaulj, Commented Oct 1, 2020 at 22:33

Tou You · Accepted Answer · 2020-10-01 23:30:05Z

1

When you give a list of np.array as input to fit() method, is considered as your model has multiple inputs.Yes, you have to create a np.array or tf.tensor from the listX. It is a good thing to convert your data to tf.data.dataset :

data_set = tf.data.Dataset.from_tensor_slices(  (your_converted_listX ,your_converted_listY) )
model.fit(data_set)

if the data is too large ,consider using tf.data.Dataset.from_generator instead.

edited Oct 1, 2020 at 23:30

answered Oct 1, 2020 at 23:19

Tou You

1,2041 gold badge9 silver badges9 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Failed to find data adapter that can handle input:...(list to numpy array with given dimensions)

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related