32930 Boost network available

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

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
dkappe
Posts: 1028
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

32930 Boost network available

Post by dkappe » Tue Apr 09, 2019 3:02 am

https://github.com/dkappe/leela-chess-w ... t-Networks

Based on my hypothesis on what was the critical element in the success of Antifish, this is 32930 sharpened with CCRL games at a low learning rate and a very few steps. Seems to give around 30 elo extra than the original net. I'll be doing this for a T40 net as well.

Javier Ros
Posts: 198
Joined: Fri Oct 12, 2012 10:48 am
Location: Seville (SPAIN)
Full name: Javier Ros

Re: 32930 Boost network available

Post by Javier Ros » Tue Apr 09, 2019 5:59 pm

Very good idea! Thanks for share!
I am testing Lc0 Boost 32930, Lc0_32930 and Lc0_41871, all with 0.21.1, on Rtx2070 against Stockfish dev. i7 4790S 8 cores
The love relationship between a chess engine tester and his computer can be summarized in one sentence:
Until heat do us part.

sovaz1997
Posts: 236
Joined: Sun Nov 13, 2016 9:37 am

Re: 32930 Boost network available

Post by sovaz1997 » Tue Apr 09, 2019 6:24 pm

I don't think that this test can talk about anything. Need more games:

Code: Select all

# PLAYER              :  RATING  ERROR  POINTS  PLAYED   (%)  CFS(%)    W    D    L  D(%)
1 32930-boost-1000    :       0     30    33.5      60  55.8      77   12   43    5  71.7
2 32930               :     -25     49    14.0      30  46.7      78    3   22    5  73.3
3 11258               :     -62     50    12.5      30  41.7     ---    2   21    7  70.0

White advantage = 78.53 +/- 24.80
Draw rate (equal opponents) = 82.20 % +/- 7.64
5-3 is not 30 Elo. It is unknown.
Zevra 2 - my chess engine.
Binary, source and description here: https://github.com/sovaz1997/Zevra2

dkappe
Posts: 1028
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

Re: 32930 Boost network available

Post by dkappe » Tue Apr 09, 2019 6:57 pm

I applied this to my own 256x20-se net called Maddex and ran 2+2 for 800 games and got +30 elo. Right now I’m running tests for 500 to 4500 step nets. 32930 may need more steps. TBD.

Javier Ros
Posts: 198
Joined: Fri Oct 12, 2012 10:48 am
Location: Seville (SPAIN)
Full name: Javier Ros

Re: 32930 Boost network available

Post by Javier Ros » Thu Apr 11, 2019 4:33 pm

Test with Silver suite 50x2 game each match:
Lc0 Boost 32930, Lc0_32930 and Lc0_41871, all with 0.21.1 version, on Rtx2070 against Stockfish dev. i7 4790S at 3.6 MGhz. 8 cores at 5 minutes plus 3 seconds under Fritz gui. First half of the test finished with 150 games played.

Lc0 Boost 32930 scored much better than 32930 and almost as good as 41871.

Lc0_41871SFdev, Blitz 5m+3s-3 2019

Stockfish 090419 64 BMI2 - Lc0 v0.21.1_41871 22.5 - 27.5 +6/=33/-11 45.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1Boost32930 23.0 - 27.0 +6/=34/-10 46.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1_32930 26.5 - 23.5 +9/=35/-6 53.00%

Program Elo + - Games Score Av.Op. Draws

1 Lc0 v0.21.1_41871 : 2428 57 55 50 55.0 % 2393 66.0 %
2 Lc0 v0.21.1Boost32930 : 2421 56 54 50 54.0 % 2393 68.0 %
3 Stockfish 090419 64 BMI2 : 2393 31 32 150 48.0 % 2407 68.0 %
4 Lc0 v0.21.1_32930 : 2372 52 54 50 47.0 % 2393 70.0 %
The love relationship between a chess engine tester and his computer can be summarized in one sentence:
Until heat do us part.

Raphexon
Posts: 423
Joined: Sun Mar 17, 2019 11:00 am
Full name: Henk Drost

Re: 32930 Boost network available

Post by Raphexon » Sat Apr 13, 2019 1:31 pm

Would it be possible to distill the latest network to 16x2?

I assume you need an Nvidia GPU for it. (Which I don't have)

Javier Ros
Posts: 198
Joined: Fri Oct 12, 2012 10:48 am
Location: Seville (SPAIN)
Full name: Javier Ros

Re: 32930 Boost network available

Post by Javier Ros » Mon Apr 15, 2019 3:07 pm

Javier Ros wrote:
Thu Apr 11, 2019 4:33 pm
Test with Silver suite 50x2 game each match:
Lc0 Boost 32930, Lc0_32930 and Lc0_41871, all with 0.21.1 version, on Rtx2070 against Stockfish dev. i7 4790S at 3.6 MGhz. 8 cores at 5 minutes plus 3 seconds under Fritz gui. First half of the test finished with 150 games played.

Lc0 Boost 32930 scored much better than 32930 and almost as good as 41871.

Lc0_41871SFdev, Blitz 5m+3s-3 2019

Stockfish 090419 64 BMI2 - Lc0 v0.21.1_41871 22.5 - 27.5 +6/=33/-11 45.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1Boost32930 23.0 - 27.0 +6/=34/-10 46.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1_32930 26.5 - 23.5 +9/=35/-6 53.00%

Program Elo + - Games Score Av.Op. Draws

1 Lc0 v0.21.1_41871 : 2428 57 55 50 55.0 % 2393 66.0 %
2 Lc0 v0.21.1Boost32930 : 2421 56 54 50 54.0 % 2393 68.0 %
3 Stockfish 090419 64 BMI2 : 2393 31 32 150 48.0 % 2407 68.0 %
4 Lc0 v0.21.1_32930 : 2372 52 54 50 47.0 % 2393 70.0 %
In the second half of the test, the results were

Lc0_41871SFdev, Blitz 5m+3s 2019

Stockfish 090419 64 BMI2 - Lc0 v0.21.1_41871 24.0 - 26.0 +4/=40/-6 48.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1Boost32930 22.5 - 27.5 +4/=37/-9 45.00%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1_32930 25.0 - 25.0 +7/=36/-7 50.00%

so that the total test result is

Lc0_41871SFdev, Blitz 5m+3s-3 2019

Stockfish 090419 64 BMI2 - Lc0 v0.21.1_41871 46.5 - 53.5 +10/=73/-17 46.50%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1Boost32930 45.5 - 54.5 +10/=71/-19 45.50%
Stockfish 090419 64 BMI2 - Lc0 v0.21.1_32930 51.5 - 48.5 +16/=71/-13 51.50%

and Lc0 with 32930 Boost gets the best result!
6 men syzygy tablebases for both.
The love relationship between a chess engine tester and his computer can be summarized in one sentence:
Until heat do us part.

Modern Times
Posts: 2893
Joined: Thu Jun 07, 2012 9:02 pm

Re: 32930 Boost network available

Post by Modern Times » Tue Apr 16, 2019 5:53 am

dkappe wrote:
Tue Apr 09, 2019 3:02 am
https://github.com/dkappe/leela-chess-w ... t-Networks

Based on my hypothesis on what was the critical element in the success of Antifish, this is 32930 sharpened with CCRL games at a low learning rate and a very few steps. Seems to give around 30 elo extra than the original net. I'll be doing this for a T40 net as well.
Which T40 net do you think you'll choose ?
.

Opinions expressed here are my own, and not necessarily those of the CCRL Group.

glennsamuel32
Posts: 128
Joined: Sat Dec 04, 2010 4:31 pm
Location: 223

Re: 32930 Boost network available

Post by glennsamuel32 » Tue Apr 16, 2019 7:19 am

Raphexon wrote:
Sat Apr 13, 2019 1:31 pm
Would it be possible to distill the latest network to 16x2?
The distillation process doesn't work on test 40 and above networks, yet.
Judge without bias, or don't judge at all...

dkappe
Posts: 1028
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

Re: 32930 Boost network available

Post by dkappe » Wed Apr 17, 2019 9:10 pm

After some additional testing, the boost net at 7000 steps is about the same head 2 head and vs sf10 at 1+1. The style of play is a bit different, however. Would appreciate feedback. https://github.com/dkappe/leela-chess-w ... t-Networks

Post Reply