Since you’re here...

We hope you will consider supporting us today. We need your support to continue to exist, because good entries are more and more work time. Every reader contribution, however big or small, is so valuable. Support "Chess Engines Diary" even a small amount– and it only takes a minute. Thank you.
============================== My email: jotes@go2.pl



Chess engine: Little Goliath 3.17 NNUE

-

 


Little Goliath - Neural Network Version, UCI chess engine.
Author: Michael Borgstädt  Rating CEDR=3243

Little Goliath ("Evolution") - Neural Network Version
The engine uses a neural network for positional evaluation. In direct comparison with the old engine, which uses manually generated evaluation functions, this engine achieves almost 100% in over 1000 test games (only a few draw games). This is obviously significantly more than other programmes have achieved when switching to neural networks. 
However, it does not seem to be due to the net itself but to its effect on a pruning system, which I call "path finding pruning" and which seems to depend essentially on an exact position evaluation. If this pruning is deactivated in both engines, the NN version "only" wins with about 70%. If I leave the pruning activated in the classic engine and play against the same version without it, the version with pruning is only just ahead (approx. 52%). 
On fast hardware, this also corresponds to the results from about 16 years ago. Consequently, the neural network or the more exact positional evaluation seems to have an extreme effect here. 
This is an important approach for future programme versions, perhaps much more can be achieved by tuning (also other techniques). Another effect is the significantly lower search depth overall. Here, too, the network seems to have an effect, the search is shallower, but more accurate. Again, it will be an approach to pruning even more aggressively. I always thought the pruning in Little Goliath was very aggressive. 
However, when I look at Stockfish, it doesn't come close to being aggressive either. I rejected similar approaches at the time, because as a chess player I was of the opinion that tactical variants are of decisive importance for the strength of the game and that the pruning here should not be too intensive. Obviously a misjudgement, which I will look at more closely. In addition, "Sigular Extensions" are now activated by default. 
This was not the case before. Here, too, the net seems to have a positive effect. Since I am often asked, here is the information that I use singular extensions in the classic form. That is, as described by Hsu with regard to Deep Blue. 
I have only changed the conditions and considerably reduced the effort for testing for singularity (at the expense of accuracy). The method of controlling singular extensions via the hash tables, which is apparently used as "state-of-the-art" today, was already tested by me 20 years ago in many programme versions without success. 
Even today, with a 1:1 implementation as in Stockfish, I cannot achieve any improvement with Little Goliath (the opposite is rather the case). Why this is so will certainly be the subject of further investigation. Apart from that, I have not made any changes. As I said, my initial aim is to highlight the differences between the two versions (classic and NN). Unfortunately, that's why all the bugs (there were some back then, I didn't keep track) of the old version (from 2006) are still included then. The engine is available as a 32-bit version, which should also run on old hardware. There is also a 64-bit version that is about 25% faster. 
 A completely revised version will follow in a few weeks, then also with the extended networks. I am personally very curious to see what will be possible here.

v.3.17:
 Adjustment of some parameters for better use of the networks for extension and pruning decisions.
Fixed some bugs, including a rare memory leak (only occurred with certain CPU/chipset combinations).
First attempts with multiprocessor support (more effort is needed here). 30 years ago this was not an issue, the subsequent adaptation is more elaborate than I thought.  
I have adapted my old bitboard code (formerly for 32-bit in "Goliath Pro"), but am dissatisfied with the speed (nps). Further optimisations are needed here, so this version still runs here with the old design. Accordingly, the pattern recognition is not yet active. These points are on the roadmap for version 4.0


Individual statistics: Little Goliath 3.15.4 - 129 games (+ 71,= 37,- 21), 69.4 %

Andscacs 0.95092              :   2 (+  1,=  1,-  0), 75.0 %
Critter 1.6a                  :   2 (+  0,=  1,-  1), 25.0 %
Vajolet2 2.8                  :   2 (+  0,=  0,-  2),  0.0 %
Fizbo 2                       :   2 (+  0,=  0,-  2),  0.0 %
Deuterium 2019.2.37.73        :   2 (+  1,=  1,-  0), 75.0 %
Dimitri 5.2.3                 :   1 (+  1,=  0,-  0), 100.0 %
Fritz 17                      :   2 (+  0,=  1,-  1), 25.0 %
SeeChess 1.2                  :   2 (+  2,=  0,-  0), 100.0 %
Jinx 1.0                      :   2 (+  2,=  0,-  0), 100.0 %
Asymptote 0.8                 :   1 (+  0,=  0,-  1),  0.0 %
NoC 4.20                      :   2 (+  2,=  0,-  0), 100.0 %
Badchess 0.4.4                :   2 (+  2,=  0,-  0), 100.0 %
Devel 4.0.2.3                 :   1 (+  0,=  1,-  0), 50.0 %
Lozza 2.2                     :   2 (+  2,=  0,-  0), 100.0 %
Barbarossa 0.6.0              :   2 (+  2,=  0,-  0), 100.0 %
Zahak 10.0                    :   2 (+  0,=  1,-  1), 25.0 %
Hiarcs 15.5                   :   4 (+  0,=  4,-  0), 50.0 %
Admete 1.5.0                  :   2 (+  2,=  0,-  0), 100.0 %
Expositor 2WQ23               :   2 (+  1,=  1,-  0), 75.0 %
Protej 0.6.2                  :   2 (+  1,=  0,-  1), 50.0 %
CT800 1.43                    :   2 (+  2,=  0,-  0), 100.0 %
FoxSEE 7.35.0                 :   2 (+  2,=  0,-  0), 100.0 %
Waxman 2020                   :   2 (+  2,=  0,-  0), 100.0 %
Loki 3.5.0                    :   2 (+  2,=  0,-  0), 100.0 %
Galjoen 0.41.2                :   2 (+  2,=  0,-  0), 100.0 %
MagiC 2                       :   1 (+  1,=  0,-  0), 100.0 %
Cosette 5.1                   :   3 (+  3,=  0,-  0), 100.0 %
Ceechess 1.4                  :   2 (+  2,=  0,-  0), 100.0 %
MinimalChess 0.6              :   2 (+  2,=  0,-  0), 100.0 %
FrankyGo 1.0.3                :   2 (+  2,=  0,-  0), 100.0 %
Blunder 7.6                   :   2 (+  2,=  0,-  0), 100.0 %
Eubos 2.10                    :   2 (+  2,=  0,-  0), 100.0 %
Tinman 0.41 dev               :   2 (+  2,=  0,-  0), 100.0 %
ChessKISS 1.8                 :   2 (+  2,=  0,-  0), 100.0 %
Cinnamon 2.4                  :   2 (+  2,=  0,-  0), 100.0 %
Hopper 1.7                    :   2 (+  2,=  0,-  0), 100.0 %
Pawn 2021-09-22               :   2 (+  2,=  0,-  0), 100.0 %
Gunborg 1.65                  :   2 (+  2,=  0,-  0), 100.0 %
Xoron b196                    :   2 (+  2,=  0,-  0), 100.0 %
Coiled 1.1                    :   2 (+  0,=  0,-  2),  0.0 %
Combusken 2.0.0               :   2 (+  0,=  1,-  1), 25.0 %
Slowchess 2.9                 :   2 (+  0,=  1,-  1), 25.0 %
Protej 0.6.3                  :   2 (+  1,=  1,-  0), 75.0 %
Journeyman 2.1                :   1 (+  1,=  0,-  0), 100.0 %
Nirvanachess 2.5              :   2 (+  1,=  1,-  0), 75.0 %
Demolito 2021-10-21           :   2 (+  0,=  2,-  0), 50.0 %
Beef 0.3.6                    :   1 (+  1,=  0,-  0), 100.0 %
FabChess 1.16                 :   2 (+  0,=  2,-  0), 50.0 %
Gargantua 1.0                 :   2 (+  0,=  2,-  0), 50.0 %
Odonata 0.6.2                 :   2 (+  1,=  1,-  0), 75.0 %
ISA 2.0.83                    :   2 (+  2,=  0,-  0), 100.0 %
Smallbrain 5.0                :   1 (+  0,=  1,-  0), 50.0 %
Dumb 1.11                     :   1 (+  0,=  1,-  0), 50.0 %
Velvet 4.1.0                  :   2 (+  0,=  2,-  0), 50.0 %
Devre 4.0                     :   2 (+  0,=  1,-  1), 25.0 %
Revenge 3.0                   :   2 (+  0,=  0,-  2),  0.0 %
Alexandria 2.4                :   1 (+  1,=  0,-  0), 100.0 %
Fornax 3.0                    :   1 (+  1,=  0,-  0), 100.0 %
Weiss 2.1 dev                 :   2 (+  0,=  2,-  0), 50.0 %
StockNemo 5.6.0.0             :   1 (+  1,=  0,-  0), 100.0 %
StockNemo 5.7.0.0             :   2 (+  0,=  1,-  1), 25.0 %
Marvin 6.1.0                  :   2 (+  1,=  1,-  0), 75.0 %
Caissa 1.1                    :   2 (+  0,=  2,-  0), 50.0 %
Frozenight 5.1.0              :   2 (+  2,=  0,-  0), 100.0 %
Caissa 1.3                    :   2 (+  0,=  1,-  1), 25.0 %
Alexandria 3.0.2              :   2 (+  0,=  2,-  0), 50.0 %
Mayhem 7.7                    :   1 (+  1,=  0,-  0), 100.0 %
RubiChess 20221120            :   2 (+  0,=  1,-  1), 25.0 %
Seer 2.4.0                    :   2 (+  0,=  0,-  2),  0.0 %


Comments

Post a Comment

Popular posts from this blog

New Strong Engines Test, by Chess Engines Diary, 2024.04.12

-
Image
  💾  552 (!) games from the tournament download   👍 @chessenginesdiary  Country - Poland, City -  Malbork 🕓 Time 3'+3" 💻HP Pavilion i5-1035G1 8GB RAM 🖬 GUI-Banksia All  CEDR  317.321 games download  (01.04.2024 - 3'+3")  HypnoS 030424, Yuli GM Pro 16, Stockfish 16.1 and ShashChess 35 . These four engines scored the same number of points and placed at the top of the table in this strong tournament. Tech table: Engine KN/move NPS dep/mov time/mov mov/game time/game fails Alexandria 6.1.0 3218 576429 30.2 5.6 56.7 316.2 Berserk 13 4779 841025 36.7 5.7 56.0 318.1 Brainlearn 28 2275 366977 31.4 6.2 43.5 269.4 Caissa 1.18 5492 923486 30.2 5.9 50.4 299.7 Clover 6.1.19 4803 854377 31.4 5.6 62.2 349.4 Cool Iris 12.10 2201 356730 27.6 6.2 51.2 315.5 CorChess 20240331 2253 372774 27.4 6.0 52.2 315.8 Crystal 8 3582 571994 20.3 6.3 47.7 298.8 Fire 9.3 5225 876565 22.0 6.0 55.5 331.1 Fisherov chess monk 1.2 3487 619728 36.0 5.6 60.4 340.0 Hyp
Read more

New chess opening book: M11.2 (bin and ctg)

-
Image
  We provide two opening books, in two different formats: bin and ctg. This is enough to be used in almost any chess program (GUI). Of course, every book may contain errors, but this one is based on a high rating. Happy testing! Book - CTG format Customers who wish to use the CTG format book will need a GUI capable of reading this format, ie. Chessbase GUIs.  Please save the book in your book folder. We strongly recommend that you "Optimize" the book in the Book options dialog so the tournament moves are chosen. The book is very wide even in tournament mode unlike most other books. A type of book format mainly used by chess programs from Chessbase , such as Fritz, Chess Base, Komodo, FatFritz, Houdini , etc. Book - BIN format (  Used by chess GUIs Arena, Banksia and Droidfish) Binary books have some important advantages, especially for computer chess engines: space-efficient, fast on accessing and searching, more useful information.  At the most simple form, they are just a
Read more

New version chess engine: Lc0 BT4

-
Image
LcZero - CEDR Rating=3683 vBT4: Big transformer 4. New network architecture which builds off of BT3 by adding two types of auxiliary heads, future heads and categorical value heads. The categorical value heads predict a distribution over values of q rather than a WDL outcome distribution, and the future heads predict the moves that will be played over the next two plies. The hope is that these heads will give additional information to the net to improve training speed. We've also fixed half-precision training, so this model will be larger. BT4 training started in mid-October and is expected to take a few months. It has 15 layers with 1024 embedding size, 32 heads per layer, and 1536 emb size, for roughly a doubling in size over BT3.  Individual statistics:  Lc0 0.30.0 Stockfish 16 5/13 -3 13 Games Booot 7.2 6/10 +2 10 Games Critter 1.6a 8/9 +7 9 Games Dragon 3.2 4/8 +0 8 Games Stockfish 20230729 3.5/8 -1 8 Games ShashChess 33.2 3.5/8 -1 8 Games Fisherov chess monk 1.2 3/
Read more

Dragon NNUE by Komodo Chess - it's free!

-
Image
    Dragon by Komodo Chess Larry Kaufman: We have released Dragon at komodochess.com. Dragon uses NNUE (Neural Network Updated Efficiently) technology, originally developed for the game of shogi. Komodo has a great deal of chess knowledge in its evaluation. Training an NNUE network based on this evaluation was both an advantage and a challenge, requiring experimentation with architectures and data generation of billions of positions. The reinforcement learning phase for Dragon is in its infancy, but is already showing great promise. Dragon is a huge strength improvement over 14.1, the last release, about 197 elo in MCTS mode and 189 elo in standard mode, at CCRL blitz time control on one thread based on direct matches. With four threads, the gains were 156 elo in MCTS and 170 in standard mode. The improvement is due to the embedded neural network providing much more accurate evaluation and also in some sense gaining an extra ply or so of search by “seeing” some tactics that a normal ev
Read more