Closed roadtorajendra closed 4 years ago
me too,do u resolve this question?
The issue was not resolved in the latest commit, instead I used the previous commit which works without this error
which code? can u send the link to me? thank you
It should be ok by now :)
which code? can u send the link to me? thank you
I think I used 2018 commintted code. Now Dany has fixed the issue for the latest release.
InvalidArgumentError (see above for traceback): Found Inf or NaN global norm. : Tensor had Inf values
I've been getting this error in the latest commit making the loss absurdly negative number or Nan. I tested with the previous versions the computation of loss is fine.