Doubt in the video

at 3:15 in video , we are creating fake data, ie noise, but shouldn’t the size of noise be equal to that of real data while passing it to discriminator , shouldn’t the size of both the real and the fake data be equal.?

Hi @saksham_thukral,

The shape of noise should not be equal to the real data. Instead, the shape of the fake images generated after passing the noise through the Generator should be equal to the real data.

Hope this helps!

ok ok got that :slight_smile: , thanks for the help