Never thought that Alliestein was in the same level with LCZero and Stockfish !

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
mwyoung
Posts: 2057
Joined: Wed May 12, 2010 8:00 pm

Re: Never thought that Alliestein was in the same level with LCZero and Stockfish !

Post by mwyoung » Tue Sep 24, 2019 5:10 pm

corres wrote:
Tue Sep 24, 2019 8:53 am
In general the chess power of an NN type engine is determined by the used NET. Particularly if the policy head - what helps engine during search - is in the given NET and this is case of Leela and her derivatives.
So if Allie as a program made for utilizing of Leela`s NET works well her chess power will be the same as Leela has.
Because a lot of important parts of Leela is in Allie (see statements of crem) for a good programmer is not a heavy task to write a good program what more or less differs from Leela and gives the same chess power.
Yes, I found little difference in the two. Regardless of what the programs report.

I also found it is the NN that makes the difference. And it makes little difference if the NN is run by Allie.exe or Lc0.exe.
Professing themselves to be wise, they became fools,
take on me. Foes 0.

Chessqueen
Posts: 645
Joined: Wed Sep 05, 2018 12:16 am
Full name: Nancy M Pichardo

Re: Never thought that Alliestein was in the same level with LCZero and Stockfish !

Post by Chessqueen » Wed Sep 25, 2019 7:12 pm

zullil wrote:
Fri Sep 13, 2019 10:16 am
corres wrote:
Fri Sep 13, 2019 9:06 am
Graham Banks wrote:
Fri Sep 13, 2019 1:39 am
Even more shocking are the two crashes of Stockfish due to a faulty file.
One more crash and it's out.

The participant authors took a vote on whether or not the SF people should be allowed to replace that file. but because one voted no, it won't happen.
Which file of Stockfish was faulty?
After many tests this is really a surprise.
The faulty file was a Windows dynamic library file: libwinpthread-1.dll. Due to insufficient testing and insufficient attention to prior bug reports, this buggy file was part of the package sent to TCEC by the Stockfish "team". The file itself is not part of the Stockfish code. At least, this is my understanding. So one more crash/hang by Stockfish and the engine is disqualified.
But now for the Superfinal, Stockfish can replace the faulty file libwinpthread-1.dll. :mrgreen:

Chessqueen
Posts: 645
Joined: Wed Sep 05, 2018 12:16 am
Full name: Nancy M Pichardo

Re: Never thought that Alliestein was in the same level with LCZero and Stockfish !

Post by Chessqueen » Wed Sep 25, 2019 7:15 pm

Jouni wrote:
Mon Sep 23, 2019 8:27 pm
SF wins league even with 2 crashes! What happened to NN revolution? BTW both SF crashes did happen when there were 20 or 40 moves with same evalution. Maybe this gives a hint about the reason?
The reason why Stockfish won is because Alliestein which is very similar to Leela took some important games away from Leela, leaving Stockfish with a greater chance to beat Leela this time. :mrgreen: :shock:

Chessqueen
Posts: 645
Joined: Wed Sep 05, 2018 12:16 am
Full name: Nancy M Pichardo

Re: Never thought that Alliestein was in the same level with LCZero and Stockfish !

Post by Chessqueen » Wed Sep 25, 2019 7:23 pm

corres wrote:
Tue Sep 24, 2019 8:53 am
In general the chess power of an NN type engine is determined by the used NET. Particularly if the policy head - what helps engine during search - is in the given NET and this is case of Leela and her derivatives.
So if Allie as a program made for utilizing of Leela`s NET works well her chess power will be the same as Leela has.
Because a lot of important parts of Leela is in Allie (see statements of crem) for a good programmer is not a heavy task to write a good program what more or less differs from Leela and gives the same chess power.
Yes you are correct the NET is very important since there was not a major change in Alliestein version except for a different an improved NET. It would be interesting to see a clash of NN engines using the same NET for all and see which NN engine is really the best.

Note: Now that Chessfighter has improved it woud be interesting if somebody here pit all these NN engine using the same NET 1: Leela, Alliestein, Stoofvlees, Scorpio and Chessfighter. By the way where is the NN version of Komodo?

mwyoung
Posts: 2057
Joined: Wed May 12, 2010 8:00 pm

Re: Never thought that Alliestein was in the same level with LCZero and Stockfish !

Post by mwyoung » Thu Sep 26, 2019 12:15 am

Chessqueen wrote:
Wed Sep 25, 2019 7:23 pm
corres wrote:
Tue Sep 24, 2019 8:53 am
In general the chess power of an NN type engine is determined by the used NET. Particularly if the policy head - what helps engine during search - is in the given NET and this is case of Leela and her derivatives.
So if Allie as a program made for utilizing of Leela`s NET works well her chess power will be the same as Leela has.
Because a lot of important parts of Leela is in Allie (see statements of crem) for a good programmer is not a heavy task to write a good program what more or less differs from Leela and gives the same chess power.
Yes you are correct the NET is very important since there was not a major change in Alliestein version except for a different an improved NET. It would be interesting to see a clash of NN engines using the same NET for all and see which NN engine is really the best.

Note: Now that Chessfighter has improved it woud be interesting if somebody here pit all these NN engine using the same NET 1: Leela, Alliestein, Stoofvlees, Scorpio and Chessfighter. By the way where is the NN version of Komodo?
Allie v0.5 + 320x24.J13B.2-swa-178000 vs. Lc0 + 320x24.J13B.2-swa-178000 5m+5s

DESKTOP-CORSAIR, Blitz 5m+5s 0

1 Lc0 v0.22.0 +15 +3/=67/-0 52.14% 36.5/70
2 Allie v0.5-dev -15 +0/=67/-3 47.86% 33.5/70
Professing themselves to be wise, they became fools,
take on me. Foes 0.

Post Reply