search

xdxuyang / Deep-Spectral-Clustering-using-Dual-Autoencoder-Network

MIT License
53 stars 17 forks source link

RuntimeError: Attempting to capture an EagerTensor without building a function. #2

Open ThibaultGROUEIX opened 4 years ago

ThibaultGROUEIX commented 4 years ago

Hi @xdxuyang, Thanks for this great work. I am going to try to run a comparison with your method.

Trying to run your method from scratch, i had this error:

RuntimeError: Attempting to capture an EagerTensor without building a function..

I've fixed this by replacing this line with self.P = tf.eye(self.x.get_shape().as_list()[1]) and adding import numpy as np Let me know if it's ok, BEst regards, Thibault

ThibaultGROUEIX commented 4 years ago

There many other bug after that, it seems related to tf version. I reverted to tf==1.5 and it runs (after fixing some paths.)

xdxuyang commented 4 years ago

Hi,

I think this may be related to tf version, and I recommend using version 1.2.

Best regards,

Xu

ThibaultGROUEIX notifications@github.com 于2020年5月16日周六 下午9:47写道：

There many other bug after that, it seems related to tf version. I reverted to tf==1.5

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/xdxuyang/Deep-Spectral-Clustering-using-Dual-Autoencoder-Network/issues/2#issuecomment-629648416, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJI2VRKPXZYOMWIT6E7J2HTRR2KOBANCNFSM4NC5OOLQ .

ThibaultGROUEIX commented 4 years ago

Thanks for the feedback :) You may want to mention it somewhere in the readme. If you have a few minutes to chat in person, I could really use your help to understand the code better (if so you can call me by skype at thibault.groueix , Sincerely