About Obsidian

[Frank Quisinsky]

Hi Graham,

what is so interesting on Discord?
Never used it and I have not the opinion that I am missing something.
Only robs you of time unnecessarily!!

All the unnecessary information, also in our media world, isn't good!
Same with What's App, TikTok, Facebook, Instagram and all such things.

Best
Frank

[Frank Quisinsky]

Hi Graham,

what is so interesting on Discord?
Never used it and I have not the opinion that I am missing something.
Only robs you of time unnecessarily!!

All the unnecessary information, also in our media world, isn't good!
Same with What's App, TikTok, Facebook, Instagram and all such things.

I also dislike chess server.
A long time I avoid that for Arena, but end of the day Martin do it.
This fight I lost for many years!

Much more interesting is the good and old chess-club!

Best
Frank

[Graham Banks]

Hi Graham,

what is so interesting on Discord?
Never used it and I have not the opinion that I am missing something.
Only robs you of time unnecessarily!!

All the unnecessary information, also in our media world, isn't good!
Same with What's App, TikTok, Facebook, Instagram and all such things.

Best
Frank

The computer chess channels on discord are where most of the programmers, plus interested others hide out.
Much busier and informative than any computer chess forums.

[Frank Quisinsky]

But all the programmers have been using TalkChess for many years.

This is the main problem, Graham.
With more and more of such things more and more special groups are formed and people interested in our hobby have to search for information on x such sites. All this takes up a lot of our spare time. Who are all the children when I walk. For 20-30 years they play with others in the fresh air, what we do in the past.

Modern things that nobody needs.

On the other hand, this forum is not easy to use.
Often not available, often I have to give my password several times before I can write here.

[Ciekce]

But all the programmers have been using TalkChess for many years.

literally not even close to "all"

99% of the relevant engine dev discussion that happens in the world, happens in the stockfish, engine programming, openbench and lc0 discords; talkchess is obsolete for that purpose

like most forums, it is a relic from a previous era of the internet

[JacquesRW]

Modern things that nobody needs.

I think we do need the Discord servers: TalkChess and CPW are full of outdated, poorly tested or just plain incorrect information regarding engine dev, not to mention a distinct lack of community here compared to, for example, the Engine Programming server.

Your unwillingness to pick up "new" technology is your problem, not ours.

[Ras]

Are we talking about Obsidian 10 from https://github.com/gab8192/Obsidian ?

As example, I compared one of the most important parts (search.cpp) with SF16. This is pretty different and not a copy. There are ideas from SF, but that's common practice and also acknowledged. Also the engine file sizes are different because Obsidian uses a small net compared to SF16, and that alone makes it very unlikely that they come up with the exakt same PV and eval. I suspect that Chessqueen's test is garbage, probably running two times the same engine, not SF16 and Obsidian 10.

I built both SF16 and Obsidian 10 from source (Linux) and gave the test position after move 14 with depth 20. Both SF16 and Obsidian at 256MB hash, single thread. The output is completely different, they don't even agree on the best move within the given depth. Despite much higher NPS, probably due to the smaller net, Obsidian takes much longer to reach depth 20.

SF16:

Code: Select all

setoption name hash value 256
position fen r1b1rbk1/1ppq2pp/2n2p2/p2np3/8/PP1P1NP1/1B1NPPBP/2RQR1K1 w - - 0 1
go depth 20
info string NNUE evaluation using nn-5af11540bbfe.nnue enabled
info depth 1 seldepth 1 multipv 1 score cp 25 nodes 53 nps 53000 hashfull 0 tbhits 0 time 1 pv e2e4
info depth 2 seldepth 2 multipv 1 score cp 25 nodes 88 nps 44000 hashfull 0 tbhits 0 time 2 pv e2e4
info depth 3 seldepth 2 multipv 1 score cp 36 nodes 134 nps 67000 hashfull 0 tbhits 0 time 2 pv h2h4
info depth 4 seldepth 2 multipv 1 score cp 52 nodes 200 nps 100000 hashfull 0 tbhits 0 time 2 pv d1c2
info depth 5 seldepth 3 multipv 1 score cp 139 nodes 253 nps 126500 hashfull 0 tbhits 0 time 2 pv h2h4
info depth 6 seldepth 4 multipv 1 score cp 116 nodes 423 nps 211500 hashfull 0 tbhits 0 time 2 pv h2h4 d5e7 h4h5
info depth 7 seldepth 5 multipv 1 score cp 56 nodes 1367 nps 455666 hashfull 0 tbhits 0 time 3 pv e2e3 d7f7 d3d4
info depth 8 seldepth 5 multipv 1 score cp 101 nodes 1879 nps 469750 hashfull 0 tbhits 0 time 4 pv e2e3 d7f7 d3d4
info depth 9 seldepth 8 multipv 1 score cp 47 nodes 5790 nps 723750 hashfull 0 tbhits 0 time 8 pv e2e3 d5b6 g2f1 d7f7 d3d4
info depth 10 seldepth 12 multipv 1 score cp 42 nodes 13764 nps 917600 hashfull 0 tbhits 0 time 15 pv d2c4 a8a6 e2e4 d5b6 c4e3
info depth 11 seldepth 11 multipv 1 score cp 24 nodes 21331 nps 1122684 hashfull 0 tbhits 0 time 19 pv d2e4 d7f7 d3d4 c6d4 f3d4 e5d4 d1d4
info depth 12 seldepth 10 multipv 1 score cp 12 nodes 28862 nps 1202583 hashfull 0 tbhits 0 time 24 pv d2e4 d7f7 e2e3 d5b6 e4c5 f8c5 c1c5
info depth 13 seldepth 16 multipv 1 score cp 14 nodes 83331 nps 1436741 hashfull 1 tbhits 0 time 58 pv d1c2 g8h8 d2c4 d7f7 e2e3 c8d7 e1d1 a8d8 h2h3 d8a8
info depth 14 seldepth 18 multipv 1 score cp 7 nodes 129684 nps 1507953 hashfull 2 tbhits 0 time 86 pv d1c2 d7f7 c1a1 c8e6 e2e3 a8d8 d3d4 e5d4 b2d4 d5b6
info depth 15 seldepth 19 multipv 1 score cp 9 nodes 235451 nps 1559278 hashfull 4 tbhits 0 time 151 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e4 d5c7 d4f5 a8d8
info depth 16 seldepth 24 multipv 1 score cp 8 nodes 318699 nps 1585567 hashfull 6 tbhits 0 time 201 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e3 d5c7 g2e4 h5g6 e4g6 f7g6 c2g6
info depth 17 seldepth 23 multipv 1 score cp 10 nodes 438027 nps 1604494 hashfull 8 tbhits 0 time 273 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e3 d5c7 c2c3 c7d5 c3c1 h5g6 h2h4
info depth 18 seldepth 22 multipv 1 score cp 4 nodes 598504 nps 1621962 hashfull 11 tbhits 0 time 369 pv d1c2 d7f7 c1a1 c8g4 e2e4 d5e7 h2h3 g4e6 d3d4 e5d4 b2d4 e8d8 d4c5 e7g6 c5f8 f7f8
info depth 19 seldepth 22 multipv 1 score cp 4 nodes 699586 nps 1634546 hashfull 14 tbhits 0 time 428 pv d1c2 d7f7 c1a1 c8g4 d3d4 e5d4 f3d4 c6d4 b2d4 a8d8 d2c4 d5e7 d4b2 e7c6 g2c6 b7c6 c4a5 d8b8 e2e4
info depth 20 seldepth 21 multipv 1 score cp 11 nodes 848354 nps 1637749 hashfull 19 tbhits 0 time 518 pv d1c2 d7f7 c1a1 c8e6 d3d4 e5d4 f3d4 c6d4 b2d4 a8d8 d4c5 f8c5 c2c5 f7e7 c5a5 b7b6 a5b5 d5e3 f2e3 d8d2
bestmove d1c2 ponder d7f7

Obsidian 10:

Code: Select all

setoption name hash value 256
position fen r1b1rbk1/1ppq2pp/2n2p2/p2np3/8/PP1P1NP1/1B1NPPBP/2RQR1K1 w - - 0 1
go depth 20
info depth 1 score cp 41 nodes 65 nps 65000 tbhits 0 time 0 pv d1c2
info depth 2 score cp 105 nodes 104 nps 104000 tbhits 0 time 0 pv d1c2 a5a4
info depth 3 score cp 75 nodes 169 nps 169000 tbhits 0 time 0 pv h2h3 a5a4 b3a4
info depth 4 score cp 406 nodes 350 nps 350000 tbhits 0 time 0 pv e2e4 a5a4 e4d5 d7d5
info depth 5 score cp 406 nodes 448 nps 448000 tbhits 0 time 0 pv e2e4 a5a4 e4d5 d7d5 b3b4
info depth 6 score cp 44 nodes 2290 nps 1145000 tbhits 0 time 2 pv d2c4 c6d8 d1c2 c7c6 e2e4
info depth 7 score cp 55 nodes 5861 nps 1172200 tbhits 0 time 5 pv d2e4 d5b6 d3d4 e5d4
info depth 8 score cp 10 nodes 20894 nps 1492428 tbhits 0 time 14 pv d2e4 d7f7 e2e3 d5b6 f3d2 e8d8 d3d4 e5d4 e3d4
info depth 9 score cp 18 nodes 35680 nps 1784000 tbhits 0 time 20 pv d2e4 d7f7 d3d4 e5d4 f3d4 c6d4 d1d4 c7c6 d4d2
info depth 10 score cp 20 nodes 50816 nps 1954461 tbhits 0 time 26 pv d2e4 d7f7 e4c5 e8d8 d3d4 e5d4 f3d4 c6d4 d1d4
info depth 11 score cp 18 nodes 95562 nps 2123600 tbhits 0 time 45 pv d2e4 d7f7 e4c5 b7b6 c5a4 c8b7 f3d2 f7d7 a4c3 c6d4 g2d5 b7d5 c3d5
info depth 12 score cp 21 nodes 161102 nps 2269042 tbhits 0 time 71 pv d2e4 d7f7 e4c5 b7b6 c5e4 c8b7 e4d2 a8d8 e2e3 d8d7 d1c2 e8d8 d3d4
info depth 13 score cp 21 nodes 213414 nps 2345208 tbhits 0 time 91 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 e5d4 f3d4 c6d4 b2d4 c7c6 d1c2 b6d5 e1d1 c8g4
info depth 14 score cp 19 nodes 294118 nps 2391203 tbhits 0 time 123 pv d2e4 d7f7 e2e3 d5b6 e4c5 b6d7 c5d7 c8d7 d3d4 e5d4 e3d4 e8e1 f3e1 d7f5 d4d5 a8d8
info depth 15 score cp 30 nodes 447322 nps 2457813 tbhits 0 time 182 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 f8c5 c1c5 c8e6 f3d2 e5d4 e3d4 e6d5 g2f1 h7h6
info depth 16 score cp 30 nodes 712482 nps 2499936 tbhits 0 time 285 pv d2e4 d7f7 e2e3 c8g4 e4c5 b7b6 c5e4 d5e7 d1c2 a8d8 d3d4 e5d4 e3d4 g4e6 e4d2 e6d5 g2f1 e7c8
info depth 17 score cp 31 nodes 872651 nps 2514844 tbhits 0 time 347 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 e5d4 e3d4 f8c5 c1c5 c8e6 f3d2 e6d5 d2c4 d5g2 g1g2 a5a4 c4b6 c7b6 c5b5
info depth 18 score cp 28 nodes 1721953 nps 2524857 tbhits 0 time 682 pv d2e4 d7f7 d1c2 e8d8 e2e3 c8e6 d3d4 e5d4 e3d4 c6e7 e4c5 e6c8 f3d2 c7c6 d2c4 e7f5 c5a4
info depth 19 score cp 30 nodes 2514045 nps 2536876 tbhits 0 time 991 pv d2e4 d7f7 d1c2 d5b6 f3d2 c8e6 e4c5 f8c5 c2c5 e6d5 g2d5 b6d5 c5c4 a8d8 c4b5 d5b6 c1c5 b6d7 c5c2 d7b6 e1c1 h7h6 c2c5
info depth 20 score cp 27 nodes 3830910 nps 2540391 tbhits 0 time 1508 pv d2e4 d7f7 d1c2 e8d8 e4c5 h7h6 e1d1 d5b6 f3d2 b6d5 e2e3 g8h8 d2f3 f8c5 c2c5 c8e6 f3d2 d8d7 c5b5 d5b6 d2c4
bestmove d2e4

[RubiChess]

TalkChess/Discord comparison...

Example of what is different: In Discord trolls like Chessqueen who edit a PGN header and claim it's a different game even of a different engine would get a livetime ban.

Example of what they have in common: Well, look at the Memes channel in SF Discord (use Google first to look up what a 'Meme' is). You'll find several posts of Talkchess "discussed" there.

Have a nice day.
Andreas

[mvanthoor]

...

This engine posted by Gabor has the same main sources:

New engine Molybdenum

rn5f107s2, Germany
https://github.com/rn5f107s2/Molybdenum ... s/tag/v2.0

But the link to the engine programming discord doesn't work.

Even so, I can't believe that there are people who go: "Let's write a chess engine!" Then: "Let's hang around in the Discord and Engine Programming Channels for a few weeks!" Then, 200 commits and a fart later, another 3500 engine drops.

That just feels wrong to me. Either I'm exceedingly stupid because I can't do that, or many others are just (re)writing code without exactly understanding what it does and why. My challenge would then be: if I gave you a text editor, a compiler of your language of choice and your own knowledge and notes (so NOT the source code of your previous engine), could you do it _again_? If not, then... well... you either didn't understand, or you didn't document and forgot.

There are SO many engines that just list a bunch of features and pop out new versions every other day without even listing a changelog. Those are useless to learn from, even if open source. The same goes for "I've been writing a chess engine for the last few weeks and *pop!* here's your new 3500 Elo engine! Enjoy, bye!" And the creator was never seen again. (And may not even have been seen before.)

I ignore all of them. The only new open source engines I follow are the ones for which I can see the beginnings, that document their progress and what they are doing, and why.

[DmitriyFrosty]

Are we talking about Obsidian 10 from https://github.com/gab8192/Obsidian ?

As example, I compared one of the most important parts (search.cpp) with SF16. This is pretty different and not a copy. There are ideas from SF, but that's common practice and also acknowledged. Also the engine file sizes are different because Obsidian uses a small net compared to SF16, and that alone makes it very unlikely that they come up with the exakt same PV and eval. I suspect that Chessqueen's test is garbage, probably running two times the same engine, not SF16 and Obsidian 10.

I built both SF16 and Obsidian 10 from source (Linux) and gave the test position after move 14 with depth 20. Both SF16 and Obsidian at 256MB hash, single thread. The output is completely different, they don't even agree on the best move within the given depth. Despite much higher NPS, probably due to the smaller net, Obsidian takes much longer to reach depth 20.

SF16:

Code: Select all

setoption name hash value 256
position fen r1b1rbk1/1ppq2pp/2n2p2/p2np3/8/PP1P1NP1/1B1NPPBP/2RQR1K1 w - - 0 1
go depth 20
info string NNUE evaluation using nn-5af11540bbfe.nnue enabled
info depth 1 seldepth 1 multipv 1 score cp 25 nodes 53 nps 53000 hashfull 0 tbhits 0 time 1 pv e2e4
info depth 2 seldepth 2 multipv 1 score cp 25 nodes 88 nps 44000 hashfull 0 tbhits 0 time 2 pv e2e4
info depth 3 seldepth 2 multipv 1 score cp 36 nodes 134 nps 67000 hashfull 0 tbhits 0 time 2 pv h2h4
info depth 4 seldepth 2 multipv 1 score cp 52 nodes 200 nps 100000 hashfull 0 tbhits 0 time 2 pv d1c2
info depth 5 seldepth 3 multipv 1 score cp 139 nodes 253 nps 126500 hashfull 0 tbhits 0 time 2 pv h2h4
info depth 6 seldepth 4 multipv 1 score cp 116 nodes 423 nps 211500 hashfull 0 tbhits 0 time 2 pv h2h4 d5e7 h4h5
info depth 7 seldepth 5 multipv 1 score cp 56 nodes 1367 nps 455666 hashfull 0 tbhits 0 time 3 pv e2e3 d7f7 d3d4
info depth 8 seldepth 5 multipv 1 score cp 101 nodes 1879 nps 469750 hashfull 0 tbhits 0 time 4 pv e2e3 d7f7 d3d4
info depth 9 seldepth 8 multipv 1 score cp 47 nodes 5790 nps 723750 hashfull 0 tbhits 0 time 8 pv e2e3 d5b6 g2f1 d7f7 d3d4
info depth 10 seldepth 12 multipv 1 score cp 42 nodes 13764 nps 917600 hashfull 0 tbhits 0 time 15 pv d2c4 a8a6 e2e4 d5b6 c4e3
info depth 11 seldepth 11 multipv 1 score cp 24 nodes 21331 nps 1122684 hashfull 0 tbhits 0 time 19 pv d2e4 d7f7 d3d4 c6d4 f3d4 e5d4 d1d4
info depth 12 seldepth 10 multipv 1 score cp 12 nodes 28862 nps 1202583 hashfull 0 tbhits 0 time 24 pv d2e4 d7f7 e2e3 d5b6 e4c5 f8c5 c1c5
info depth 13 seldepth 16 multipv 1 score cp 14 nodes 83331 nps 1436741 hashfull 1 tbhits 0 time 58 pv d1c2 g8h8 d2c4 d7f7 e2e3 c8d7 e1d1 a8d8 h2h3 d8a8
info depth 14 seldepth 18 multipv 1 score cp 7 nodes 129684 nps 1507953 hashfull 2 tbhits 0 time 86 pv d1c2 d7f7 c1a1 c8e6 e2e3 a8d8 d3d4 e5d4 b2d4 d5b6
info depth 15 seldepth 19 multipv 1 score cp 9 nodes 235451 nps 1559278 hashfull 4 tbhits 0 time 151 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e4 d5c7 d4f5 a8d8
info depth 16 seldepth 24 multipv 1 score cp 8 nodes 318699 nps 1585567 hashfull 6 tbhits 0 time 201 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e3 d5c7 g2e4 h5g6 e4g6 f7g6 c2g6
info depth 17 seldepth 23 multipv 1 score cp 10 nodes 438027 nps 1604494 hashfull 8 tbhits 0 time 273 pv d1c2 d7f7 c1a1 c8g4 d3d4 c6d4 b2d4 e5d4 f3d4 g4h5 d2c4 c7c6 e2e3 d5c7 c2c3 c7d5 c3c1 h5g6 h2h4
info depth 18 seldepth 22 multipv 1 score cp 4 nodes 598504 nps 1621962 hashfull 11 tbhits 0 time 369 pv d1c2 d7f7 c1a1 c8g4 e2e4 d5e7 h2h3 g4e6 d3d4 e5d4 b2d4 e8d8 d4c5 e7g6 c5f8 f7f8
info depth 19 seldepth 22 multipv 1 score cp 4 nodes 699586 nps 1634546 hashfull 14 tbhits 0 time 428 pv d1c2 d7f7 c1a1 c8g4 d3d4 e5d4 f3d4 c6d4 b2d4 a8d8 d2c4 d5e7 d4b2 e7c6 g2c6 b7c6 c4a5 d8b8 e2e4
info depth 20 seldepth 21 multipv 1 score cp 11 nodes 848354 nps 1637749 hashfull 19 tbhits 0 time 518 pv d1c2 d7f7 c1a1 c8e6 d3d4 e5d4 f3d4 c6d4 b2d4 a8d8 d4c5 f8c5 c2c5 f7e7 c5a5 b7b6 a5b5 d5e3 f2e3 d8d2
bestmove d1c2 ponder d7f7

Obsidian 10:

Code: Select all

setoption name hash value 256
position fen r1b1rbk1/1ppq2pp/2n2p2/p2np3/8/PP1P1NP1/1B1NPPBP/2RQR1K1 w - - 0 1
go depth 20
info depth 1 score cp 41 nodes 65 nps 65000 tbhits 0 time 0 pv d1c2
info depth 2 score cp 105 nodes 104 nps 104000 tbhits 0 time 0 pv d1c2 a5a4
info depth 3 score cp 75 nodes 169 nps 169000 tbhits 0 time 0 pv h2h3 a5a4 b3a4
info depth 4 score cp 406 nodes 350 nps 350000 tbhits 0 time 0 pv e2e4 a5a4 e4d5 d7d5
info depth 5 score cp 406 nodes 448 nps 448000 tbhits 0 time 0 pv e2e4 a5a4 e4d5 d7d5 b3b4
info depth 6 score cp 44 nodes 2290 nps 1145000 tbhits 0 time 2 pv d2c4 c6d8 d1c2 c7c6 e2e4
info depth 7 score cp 55 nodes 5861 nps 1172200 tbhits 0 time 5 pv d2e4 d5b6 d3d4 e5d4
info depth 8 score cp 10 nodes 20894 nps 1492428 tbhits 0 time 14 pv d2e4 d7f7 e2e3 d5b6 f3d2 e8d8 d3d4 e5d4 e3d4
info depth 9 score cp 18 nodes 35680 nps 1784000 tbhits 0 time 20 pv d2e4 d7f7 d3d4 e5d4 f3d4 c6d4 d1d4 c7c6 d4d2
info depth 10 score cp 20 nodes 50816 nps 1954461 tbhits 0 time 26 pv d2e4 d7f7 e4c5 e8d8 d3d4 e5d4 f3d4 c6d4 d1d4
info depth 11 score cp 18 nodes 95562 nps 2123600 tbhits 0 time 45 pv d2e4 d7f7 e4c5 b7b6 c5a4 c8b7 f3d2 f7d7 a4c3 c6d4 g2d5 b7d5 c3d5
info depth 12 score cp 21 nodes 161102 nps 2269042 tbhits 0 time 71 pv d2e4 d7f7 e4c5 b7b6 c5e4 c8b7 e4d2 a8d8 e2e3 d8d7 d1c2 e8d8 d3d4
info depth 13 score cp 21 nodes 213414 nps 2345208 tbhits 0 time 91 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 e5d4 f3d4 c6d4 b2d4 c7c6 d1c2 b6d5 e1d1 c8g4
info depth 14 score cp 19 nodes 294118 nps 2391203 tbhits 0 time 123 pv d2e4 d7f7 e2e3 d5b6 e4c5 b6d7 c5d7 c8d7 d3d4 e5d4 e3d4 e8e1 f3e1 d7f5 d4d5 a8d8
info depth 15 score cp 30 nodes 447322 nps 2457813 tbhits 0 time 182 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 f8c5 c1c5 c8e6 f3d2 e5d4 e3d4 e6d5 g2f1 h7h6
info depth 16 score cp 30 nodes 712482 nps 2499936 tbhits 0 time 285 pv d2e4 d7f7 e2e3 c8g4 e4c5 b7b6 c5e4 d5e7 d1c2 a8d8 d3d4 e5d4 e3d4 g4e6 e4d2 e6d5 g2f1 e7c8
info depth 17 score cp 31 nodes 872651 nps 2514844 tbhits 0 time 347 pv d2e4 d7f7 e2e3 d5b6 e4c5 e8d8 d3d4 e5d4 e3d4 f8c5 c1c5 c8e6 f3d2 e6d5 d2c4 d5g2 g1g2 a5a4 c4b6 c7b6 c5b5
info depth 18 score cp 28 nodes 1721953 nps 2524857 tbhits 0 time 682 pv d2e4 d7f7 d1c2 e8d8 e2e3 c8e6 d3d4 e5d4 e3d4 c6e7 e4c5 e6c8 f3d2 c7c6 d2c4 e7f5 c5a4
info depth 19 score cp 30 nodes 2514045 nps 2536876 tbhits 0 time 991 pv d2e4 d7f7 d1c2 d5b6 f3d2 c8e6 e4c5 f8c5 c2c5 e6d5 g2d5 b6d5 c5c4 a8d8 c4b5 d5b6 c1c5 b6d7 c5c2 d7b6 e1c1 h7h6 c2c5
info depth 20 score cp 27 nodes 3830910 nps 2540391 tbhits 0 time 1508 pv d2e4 d7f7 d1c2 e8d8 e4c5 h7h6 e1d1 d5b6 f3d2 b6d5 e2e3 g8h8 d2f3 f8c5 c2c5 c8e6 f3d2 d8d7 c5b5 d5b6 d2c4
bestmove d2e4

You need compare it to Stockfish with pure nnue evaluation (when he uses one network)