You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Sorry to bother, I'm fairly new to Keras and everything. I'm trying to set up two adversarial neural networks, and your lib seems very interesting to me. I cloned the repository and installation went successfully. But I tried to launch example_gan.py, and I got an error. I don't understand enough yet to be able to solve it by myself...
I'm using Spyder 2.3.9, with Python 3.5, TensorFlow : 1.1.0, Keras : 2.0.4
The error is : "TypeError: Value passed to parameter 'shape' has DataType float32 not in list of allowed values: int32, int64"
Issue with python 3. division is an int in python 2 but float in python 3. Dense(hidden_dim/2) should be Dense(int(hidden_dim/2)). Just add an int cast to the line that is failing.
Thank you very much for fixing this !
I'm sorry to have not seen the PR related...
A tiny thing I changed myself, you might want to know about this :
line 110 : example_gan(AdversarialOptimizerSimultaneous(), "output/gan"),
I just changed "output/gan" to os.path.join("output", "gan"), since on Windows it was a problem.
I know it's not really important, I just wanted to help.
Hello,
Sorry to bother, I'm fairly new to Keras and everything. I'm trying to set up two adversarial neural networks, and your lib seems very interesting to me. I cloned the repository and installation went successfully. But I tried to launch example_gan.py, and I got an error. I don't understand enough yet to be able to solve it by myself...
I'm using Spyder 2.3.9, with Python 3.5, TensorFlow : 1.1.0, Keras : 2.0.4
The error is : "TypeError: Value passed to parameter 'shape' has DataType float32 not in list of allowed values: int32, int64"
PasteBin of the full Traceback : https://pastebin.com/0BaAcuQa
Any kind of help is appreciated :) !
The text was updated successfully, but these errors were encountered: