AttributeError: 'Tensor' object has no attribute 'assign' about dl-4-tsc

[hfawaz]

Hello,

I am very glad that my code and paper were helpful !!

As for your error regarding MCNN, I think the network is being fed a kernel_size equal to zero.

Can you make sure what values you are choosing for the kernel_size ?

kernel_size = int(ori_len * filter_size)

You should make sure that this equation is not zero.

In your case ori_len=14 (if I am not wrong).

Then you should change the values in filter_sizes.

To sum up, try to remove the 0.05 in this line.

Let me know if this fixes the error.

Best regards,

[tsafs]

Thank you very much for your quick response. Indeed, your proposed solution fixed the error.

Best regards,

[hfawaz]

Glad to help!

Do not hesitate if you have any other questions.

Best regards,

Hassan

[jayanti-prasad]

where is the answer ?
This is strange that 'tf' has an object 'assign' but that does not seem working !

import tensorflow.compat.v1 as tf
tf.disable_eager_execution()
theta=tf.constant(1)
gradiants=tf.constant(3)
learning_rate=tf.constant(2)
dir(tf.assign)
['annotations', 'call', 'class', 'closure', 'code', 'defaults', 'delattr', 'dict', 'dir', 'doc', 'eq', 'format', 'ge', 'get', 'getattribute', 'globals', 'gt', 'hash', 'init', 'init_subclass', 'kwdefaults', 'le', 'lt', 'module', 'name', 'ne', 'new', 'qualname', 'reduce', 'reduce_ex', 'repr', 'setattr', 'sizeof', 'str', 'subclasshook', '_tf_api_names', '_tf_api_names_v1']
training_op = tf.assign(theta, theta - learning_rate * gradiants)
Traceback (most recent call last):
File "", line 1, in
File "/usr/local/lib/python3.7/site-packages/tensorflow_core/python/ops/state_ops.py", line 228, in assign
return ref.assign(value, name=name)
AttributeError: 'Tensor' object has no attribute 'assign'