importing TGAN about tgan

[ivanguerron21]

Happened to me as well, I copied the original code and tried to use it but had a lot of problems. I am now stuck with this error ValueError: Tried to convert 't' to a tensor and failed. Error: Shapes must be equal rank, but are 2 and 1
From merging shape 5 with other shapes. for '{{node GAN_loss/discrim/L2Loss/packed}} = Pack[N=7, T=DT_FLOAT, axis=0](GAN_loss/discrim/L2Loss/packed/ReadVariableOp, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_1, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_2, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_3, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_4, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_5, GAN_loss/discrim/L2Loss/packed/ReadVariableOp_6)' with input shapes: [26,100], [100], [100,100], [100], [110], [110,1], [1].
Let me know if you come up with something for this

[AryanSaeedi]

@ ivanguerron21 I had the same problem at some point. try working with a virtual environment as suggested and use the python 3.7 or any other version suggested here https://github.com/sdv-dev/TGAN. Also make sure to install tensorflow version 1.15.0. it worked for me