Training model is only rating horses.
Hi everyone, I’m running the model from C1_W4_Lab_1_image_generator_no_validation…
I don’t know what I did wrong, but whenever I load any image, the result remains the same, classifying all loaded images as “is a horse”.
What could have happened?
Hey @menezesluiz ,
Have you tried to normalise the input, the
x, when you did prediction?
Because the training performed with the normalisation earlier , my understanding is that the sampling or prediction should do this as well.
Hopefully, it helps
Thank you Mr. Chris.X.
Sorry for the delay in responding. I was a little absent here in the community and in the formations, I was involved in other projects in parallel.