I was impressed by this article - link.
Despite only being a short article, it contains complete instructions for training TensorFlow with chess skill.
It's not perfect - but I've never seen chess skill training made this easy before!
Short Article With Complete NN Chess Training Procedure
Moderators: hgm, Rebel, chrisw
-
- Posts: 11574
- Joined: Thu Mar 09, 2006 12:57 am
- Location: Birmingham UK
Short Article With Complete NN Chess Training Procedure
Writing is the antidote to confusion.
It's not "how smart you are", it's "how are you smart".
Your brain doesn't work the way you want, so train it!
It's not "how smart you are", it's "how are you smart".
Your brain doesn't work the way you want, so train it!
-
- Posts: 97
- Joined: Mon Jun 25, 2012 10:16 pm
- Location: Forks, WA
- Full name: Ben Nye
Re: Short Article With Complete NN Chess Training Procedure
That is some seriously horrific code. And the basic idea(trying to have the network predict a move directly in algebraic notation using three outputs, x, y, and piece type) is pretty terrible also.
One part that really sticks out as do-not-copy is where he tried to one-hot encode the column with a=no bits set, b=first bit set, c=second bit set instead of a=first bit, b=second bit, c=third bit. Overall, you would be better off not learning from that.
One part that really sticks out as do-not-copy is where he tried to one-hot encode the column with a=no bits set, b=first bit set, c=second bit set instead of a=first bit, b=second bit, c=third bit. Overall, you would be better off not learning from that.