Closed DrhF closed 4 years ago
Hi @DrhF.
Thanks for your interest in our research. Your fix to the first problem was bound to give a ValueError
since the Reshape() layer shapes the output of the previous layers to have the shape (21, 22, 22, 22) (as you can see from the comment above the Reshape() line).
As for the fix to the first problem, the Reshape() layer is probably giving you the shape as (21, 22, 22, 3, None) instead of (None, 21, 22, 22, 3). Like you said, this might be due to different tensorflow and keras versions.
The versions used by me are:
tensorflow == 1.13.1
keras ==2.2.4
Hint: Do check out the data_format
parameter in the docs for the Conv3D layer concerning channels_first
and channels_last
.
Hello! I was trying to reproduce experiments with your solution. When I run train.py I have this error:
I tried to fix that commenting out this line in mymodel.py
And got an error:
Could you please specify tensorflow and keras versions? (I guess this is the issue: I use different versions of tensorflow)
If not, is there any suggestion to fix those errors?