NiCimEngine Public release!

[yodatk]

Hello everyone!

I have been working on a Chess engine for a while now, and I am excited to say that a public release of my Engine is finaly ready!
you can see the source code and download it from here:

https://github.com/yodatk/NiCimEngine

you can even try and play against it in Lichess:

https://lichess.org/@/NissimEngine
(currently can play bullet, Blitz, and Rapid)

Info On the engine:
* Written in C++, With NNUE support ( with help from this source code: https://github.com/dshawul/nnue-probe)
* Working with UCI protocol.
* Currently compiling to with Cmake (really easy with CLIon) or with makefile (make file is for windows only. sorry )
* I Test against tscp and Vice and NiCim won both pretty easily. I also tested it with an Engine called FrankWalter ( you can see it's ratings here: https://ccrl.chessdom.com/ccrl/4040/cgi ... 4_0_64-bit) and after a lot of games of bullet, blitz and Rapid - Nicim had total of 70% precent of the winings.

If someone knows anyone from CCRL, I would really like to test it over there.

THANKS A LOT to Maksym, from Chess Programming youtube channel (https://www.youtube.com/channel/UCB9-pr ... KKqDgXhsMQ) , who had an amazing chess engine toturial, and also convinced me to publish my engine.

p.s - for all the people that want to use my engine- notice that the .nnue file must be in the same directory for the engine to work properly. otherwise the engine will make silly moves....

[Graham Banks]

I'm sure that both CCRL and CEGT will test it.

[mvanthoor]

Congratulations with the release. If it scores 70% against FrankWalter (2537), the difference is about +147 Elo. Rough estimate of your engine would be tentatively 2675 (with a big error margin due to only testing against one engine). Can it play without NNUE, and if so, did you test what it scores then?

[AndrewGrant]

Does this come with your own set of weights, or are you reusing a Stockfish set?

[Guenther]

Congratulations with the release. If it scores 70% against FrankWalter (2537), the difference is about +147 Elo. Rough estimate of your engine would be tentatively 2675 (with a big error margin due to only testing against one engine). Can it play without NNUE, and if so, did you test what it scores then?

Obviously not... for the rest I better wait until I have looked at the code (see comparison vs. BBC 1.4 from cmd)...mhhh

Code: Select all

    Hash table is initialized with 2796202 entries
DELAY_TIME: 50
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
BOOK: BOOK FILE NOT FOUND
id name NiCim 3.4
id author yodatk
option name Hash type spin default 64 min 4 max 128
option name Book type check default true
uciok
uci
id name NiCim 3.4
id author yodatk
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     A B C D E F G H

     Side:     WHITE
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580

isready
readyok
go infinite
time: -1 inc:0  start: 21056812  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 16 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 63 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 125 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 187 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 375 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 577 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 905 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1545 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2387 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3900 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 7254 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 13229 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

Code: Select all

Hash table is initialied with 2796202 entries
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
uci
id name BBC  - FINAL VERSION (1.4 + SF NNUE)
id author Code Monkey King
option name Hash type spin default 64 min 4 max 128
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     a b c d e f g h

     Side:     white
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580
     Fifty move: 0

isready
readyok
go infinite
time: -1  inc: 0  start: 21181238  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 31 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 62 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 109 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 172 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 343 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 515 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 811 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1388 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2153 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3479 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 6193 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 11326 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

[yodatk]

Congratulations with the release. If it scores 70% against FrankWalter (2537), the difference is about +147 Elo. Rough estimate of your engine would be tentatively 2675 (with a big error margin due to only testing against one engine). Can it play without NNUE, and if so, did you test what it scores then?

a week ago I tested it without NNUE with my own evaluation function, to be honest I Don't remeber the results, obiusly they werent as good as with the NNUE

[yodatk]

Does this come with your own set of weights, or are you reusing a Stockfish set?

I am re-using stockfish set,
This year I am going to study my self more about NN and hopefully I will have enough knowledge to to use my own weights and training set.

[yodatk]

Congratulations with the release. If it scores 70% against FrankWalter (2537), the difference is about +147 Elo. Rough estimate of your engine would be tentatively 2675 (with a big error margin due to only testing against one engine). Can it play without NNUE, and if so, did you test what it scores then?

Obviously not... for the rest I better wait until I have looked at the code (see comparison vs. BBC 1.4 from cmd)...mhhh

Code: Select all

    Hash table is initialized with 2796202 entries
DELAY_TIME: 50
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
BOOK: BOOK FILE NOT FOUND
id name NiCim 3.4
id author yodatk
option name Hash type spin default 64 min 4 max 128
option name Book type check default true
uciok
uci
id name NiCim 3.4
id author yodatk
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     A B C D E F G H

     Side:     WHITE
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580

isready
readyok
go infinite
time: -1 inc:0  start: 21056812  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 16 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 63 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 125 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 187 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 375 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 577 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 905 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1545 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2387 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3900 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 7254 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 13229 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

Code: Select all

Hash table is initialied with 2796202 entries
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
uci
id name BBC  - FINAL VERSION (1.4 + SF NNUE)
id author Code Monkey King
option name Hash type spin default 64 min 4 max 128
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     a b c d e f g h

     Side:     white
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580
     Fifty move: 0

isready
readyok
go infinite
time: -1  inc: 0  start: 21181238  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 31 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 62 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 109 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 172 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 343 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 515 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 811 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1388 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2153 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3479 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 6193 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 11326 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

you are right - there is no plan B if it doesn't found the NNUE file,
even though in the source file you can see in Evaluation.h file my old evaluation function.
I am still trying to figure out how to toggle between them in case the NNUE file wasn't loaded properly, but I Didn't find an elegant way yet

[yodatk]

Congratulations with the release. If it scores 70% against FrankWalter (2537), the difference is about +147 Elo. Rough estimate of your engine would be tentatively 2675 (with a big error margin due to only testing against one engine). Can it play without NNUE, and if so, did you test what it scores then?

Obviously not... for the rest I better wait until I have looked at the code (see comparison vs. BBC 1.4 from cmd)...mhhh

Code: Select all

    Hash table is initialized with 2796202 entries
DELAY_TIME: 50
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
BOOK: BOOK FILE NOT FOUND
id name NiCim 3.4
id author yodatk
option name Hash type spin default 64 min 4 max 128
option name Book type check default true
uciok
uci
id name NiCim 3.4
id author yodatk
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     A B C D E F G H

     Side:     WHITE
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580

isready
readyok
go infinite
time: -1 inc:0  start: 21056812  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 16 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 63 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 125 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 187 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 375 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 577 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 905 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1545 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2387 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3900 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 7254 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 13229 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

Code: Select all

Hash table is initialied with 2796202 entries
Loading NNUE : nn-eba324f53044.nnue
NNUE file not found!
uci
id name BBC  - FINAL VERSION (1.4 + SF NNUE)
id author Code Monkey King
option name Hash type spin default 64 min 4 max 128
uciok
ucinewgame

  8  r n b q k b n r
  7  p p p p p p p p
  6  . . . . . . . .
  5  . . . . . . . .
  4  . . . . . . . .
  3  . . . . . . . .
  2  P P P P P P P P
  1  R N B Q K B N R

     a b c d e f g h

     Side:     white
     Enpassant:   no
     Castling:  KQkq

     Hash key:  6ed57b118ae99580
     Fifty move: 0

isready
readyok
go infinite
time: -1  inc: 0  start: 21181238  stop: 0  depth: 64  timeset:0
info score cp 0 depth 1 nodes 21 time 0 pv a2a3
info score cp 0 depth 2 nodes 83 time 16 pv a2a3 a7a6
info score cp 0 depth 3 nodes 213 time 31 pv a2a3 a7a6 a3a4
info score cp 0 depth 4 nodes 736 time 31 pv a2a3 a7a6 a3a4 b7b6
info score cp 0 depth 5 nodes 1362 time 47 pv a2a3 a7a6 a3a4 b7b6 a4a5
info score cp 0 depth 6 nodes 2499 time 62 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5
info score cp 0 depth 7 nodes 4554 time 78 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5
info score cp 0 depth 8 nodes 8864 time 109 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6
info score cp 0 depth 9 nodes 18364 time 172 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6
info score cp 0 depth 10 nodes 31618 time 250 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6
info score cp 0 depth 11 nodes 51945 time 343 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3
info score cp 0 depth 12 nodes 85918 time 515 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6
info score cp 0 depth 13 nodes 147754 time 811 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4
info score cp 0 depth 14 nodes 282919 time 1388 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4
info score cp 0 depth 15 nodes 456903 time 2153 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3
info score cp 0 depth 16 nodes 750991 time 3479 pv a2a3 a7a6 a3a4 b7b6 a4a5 b6a5 a1a5 c7c6 a5a6 b8a6 b2b3 d7d6 b3b4 a6b4 c2c3 e7e6
info score cp 0 depth 17 nodes 1344287 time 6193 pv b2b3 b7b6 a2a4 c7c6 c2c3 d7d6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 e7e6 d2d4 a7a6 g1f3
info score cp 0 depth 18 nodes 2489564 time 11326 pv c2c3 c7c6 b2b3 b7b6 d2d4 d7d6 a2a4 a7a6 g2g3 e7e6 a4a5 b6a5 a1a5 d8a5 b3b4 a5b4 c3b4 g8f6

you are right - there is no plan B if it doesn't found the NNUE file,
even though in the source file you can see in Evaluation.h file my old evaluation function.
I am still trying to figure out how to toggle between them in case the NNUE file wasn't loaded properly, but I Didn't find an elegant way yet

I Even try to use the old Evaluation when the engine is low on time,
sadly i did not succeed in doing so properly yet

[mvanthoor]

Did you actually _derive_ your engine from BBC?

The output is very similar; including the sending of the "id" responses at the engine start before it even receives "uci". Older BBC versions did this as well, but it seems that after I reported it, Maksim has fixed it in the version you are running.