search

PeterChe1990 / GRU-D

GRU-D, a GRU-based model with trainable decays for multivariate time series classification with missing values/irregular samplings
MIT License
118 stars 37 forks source link

Create_model Problem #5

Closed nguyendohoangkhoi closed 2 years ago

nguyendohoangkhoi commented 3 years ago

image I have an error when I run create model code. I dont understand that (None, 4) should be equal to (None, 4) ValueError: The input x and the masking m should have the same input shape, but got (None, 4) and (None, 4).

quarkmatter commented 2 years ago

Hi,

I am also getting the same error. I was able to fix it by modifying line 78 in grud_layers.py:

if input_shape[0] != input_shape[1]: --> if input_shape[0].as_list() != input_shape[1].as_list():

Yovoss commented 2 years ago

@quarkmatter Hello!,when i import all the function from tensorflow.keras rather than keras, i meet the same question! but when i take your solution, there is another error: " AttributeError: 'GRUDCell' object has no attribute 'kernel_z'" Have you meet this question? thank you!

PeterChe1990 commented 2 years ago

Please strictly follow the package dependencies in requirements.txt, as running on later versions has not been tested.