Generator param size differ

[juliecious]

Environment details

If you are already running CTGAN, please indicate the following details about the environment in which you are running it:

• CTGAN version: 0.4.3 dev2
• Python version: 3.8
• Operating System: MacOS

Problem description

I trained multiple CTGAN models with the exact same data set (demo adult income data set). After training, I tried to average the learnable parameters of the generator from each model, but they are not the same size. Sometimes it differs, sometimes it doesn't. Could you please explain why? Is it expected?

# generator params from three trained CTGANs
ValueError: operands could not be broadcast together with shapes (157,744) (158,744) (157,744)