problem in results after training

[miltosc]

Hello Memo!!!

i trained my model with 20000 photos form space to simulate a video you published zooming your eye etc The result has nothing to do with your results I would really appreciate your help. The result from a video with scarlett jojanson speech is the following. what am i doing wrong?

I run it with 500 epochs I suspect something wrong when run preproccess.py am i right? Can you please help?

Thank you in advance!!