AttributeError: 'numpy.dtype' object has no attribute 'is_floating'

Question

I develop my own op using eager execution. I met an error. How to fix it?

Env: Colaboratory TF ver.: 1.14

I use:

mnist = input_data.read_data_sets("/tmp/data/", one_hot=True)

I tried tf.convert_to_tensor() but it led to other error. Error caused by:

loss_value, grads = grad(x_, y)

Error message:

    962     flat_sources = [_handle_or_self(x) for x in flat_sources]
    963     for t in flat_sources_raw:
--> 964       if not t.dtype.is_floating:
    965         logging.vlog(
    966             logging.WARN, "The dtype of the source tensor must be "

AttributeError: 'numpy.dtype' object has no attribute 'is_floating'

mkrieger1 · Accepted Answer · 2023-01-30 12:54:57Z

1

I got that reason of the error which indicates "if not is_floating" for the raising the exception.

To solve the error, it needs to make variable to a tensor, just using convert_to_tensor() before an operation.

edited Jan 30, 2023 at 12:54

mkrieger1

24.2k7 gold badges68 silver badges84 bronze badges

answered Aug 17, 2019 at 16:11

S. Takano

3132 silver badges14 bronze badges

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

Collectives™ on Stack Overflow

AttributeError: 'numpy.dtype' object has no attribute 'is_floating'

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