Article on creation of weights for FF2

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Dann Corbit
Posts: 12482
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Article on creation of weights for FF2

Post by Dann Corbit »

Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
AndrewGrant
Posts: 1660
Joined: Tue Apr 19, 2016 6:08 am
Location: U.S.A
Full name: Andrew Grant

Re: Article on creation of weights for FF2

Post by AndrewGrant »

It has been tested in a deep match of ten thousand games against the February 11 build of Stockfish, which has since become the new Stockfish 13.
Thats the kind of thing you write when you lie about the data, and then have to protect yourself against the claim of "How could you have played 10,000 games vs Stockfish 13 so quickly?". I'de like that independent tester to make the claim, since Albert has a history of lies and deceit from trivial things to crucial ones.
Talkchess is dead without moderation. If you want my attention, contact me via andrew@grantnet.us
noobpwnftw
Posts: 560
Joined: Sun Nov 08, 2015 11:10 pm

Re: Article on creation of weights for FF2

Post by noobpwnftw »

One way to make "I put a lot of effort in translating all variable names into Dutch and it is faster and stronger now some bucks please?" sounds not like a scam, I'm buying it.
Dann Corbit
Posts: 12482
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Article on creation of weights for FF2

Post by Dann Corbit »

AndrewGrant wrote: Mon Feb 22, 2021 4:20 am
It has been tested in a deep match of ten thousand games against the February 11 build of Stockfish, which has since become the new Stockfish 13.
Thats the kind of thing you write when you lie about the data, and then have to protect yourself against the claim of "How could you have played 10,000 games vs Stockfish 13 so quickly?". I'de like that independent tester to make the claim, since Albert has a history of lies and deceit from trivial things to crucial ones.
What are some of the specific things he lied about?
I don't read all the threads here.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
NoChessNoLife
Posts: 16
Joined: Thu Feb 18, 2021 10:06 pm
Full name: Lev Chen

Re: Article on creation of weights for FF2

Post by NoChessNoLife »

AndrewGrant wrote: Mon Feb 22, 2021 4:20 am
It has been tested in a deep match of ten thousand games against the February 11 build of Stockfish, which has since become the new Stockfish 13.
Thats the kind of thing you write when you lie about the data, and then have to protect yourself against the claim of "How could you have played 10,000 games vs Stockfish 13 so quickly?". I'de like that independent tester to make the claim, since Albert has a history of lies and deceit from trivial things to crucial ones.
SF 200211 dev version = SF13. Therefore, Albert is telling the truth.
AndrewGrant
Posts: 1660
Joined: Tue Apr 19, 2016 6:08 am
Location: U.S.A
Full name: Andrew Grant

Re: Article on creation of weights for FF2

Post by AndrewGrant »

NoChessNoLife wrote: Mon Feb 22, 2021 4:38 am
AndrewGrant wrote: Mon Feb 22, 2021 4:20 am
It has been tested in a deep match of ten thousand games against the February 11 build of Stockfish, which has since become the new Stockfish 13.
Thats the kind of thing you write when you lie about the data, and then have to protect yourself against the claim of "How could you have played 10,000 games vs Stockfish 13 so quickly?". I'de like that independent tester to make the claim, since Albert has a history of lies and deceit from trivial things to crucial ones.
SF 200211 dev version = SF13. Therefore, Albert is telling the truth.
He may be telling the truth; He may not be. The matching of the dates does not prove anything; Its the detail I pointed out, not a verification of any theory.
Talkchess is dead without moderation. If you want my attention, contact me via andrew@grantnet.us
noobpwnftw
Posts: 560
Joined: Sun Nov 08, 2015 11:10 pm

Re: Article on creation of weights for FF2

Post by noobpwnftw »

NoChessNoLife wrote: Mon Feb 22, 2021 4:38 am
AndrewGrant wrote: Mon Feb 22, 2021 4:20 am
It has been tested in a deep match of ten thousand games against the February 11 build of Stockfish, which has since become the new Stockfish 13.
Thats the kind of thing you write when you lie about the data, and then have to protect yourself against the claim of "How could you have played 10,000 games vs Stockfish 13 so quickly?". I'de like that independent tester to make the claim, since Albert has a history of lies and deceit from trivial things to crucial ones.
SF 200211 dev version = SF13. Therefore, Albert is telling the truth.
No. Why? Does the Github code compiles into actual FF2?
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Article on creation of weights for FF2

Post by Ferdy »

AS probably forgot to mention an important part of how NNUE can be created.