i get 1.0 fint for all the generations

[1feres1]

I didn't make any change in the code but I get different result from the original article

[ahmedfgad]

The reason is that the dataset used can be classified easily. This is why most of the solutions give high classification accuracy. In the tutorial, the accuracy at generation 0 was 0.966 which is pretty high also.

Try to only use few training samples and you can sense the change.

I may prepare a more complicated example. But things are working properly.