Not getting same/similar results by using your code & database at Frame Prediction Visualizations section

[Marianatorres0752]

I was training the same database with author'scode but I'm not having the same or similar results at Frame Prediction Visualizations
section from https://keras.io/examples/vision/conv_lstm/ Seems that the images are getting a gray background, I'd love to see if you could support me by mentioning me if I failed in a certain part of implementing author's code :)

Here's my colab : https://colab.research.google.com/drive/1S8JN9WBwAFaSuuPd28X8697EoYHWHGfN?usp=sharing

Thanks for your efforts and all in advance :)

[sachinprasadhs]

Closing it as duplicate of #732

[github-actions]

Are you satisfied with the resolution of your issue?
Yes
No