Stockfish Development Versions are build automatically if there are changes on the master branch in the git repository (https://github.com/official-stockfish/Stockfish). Use it at your own risk.
They are compiled with gcc 11.2/mingw 10 on Ubuntu 22.04.

!! latest version !!


Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Tue May 21 22:06:17 2024 +0200
Timestamp: 1716321977

Lower smallnet threshold with updated eval divisors

Params found after 30k spsa games at 60+0.6, with initial
values from 64k spsa games at 45+0.45

First spsa with 64k / 120k games at 45+0.45:
https://tests.stockfishchess.org/tests/view/664a561b5fc7b70b8817c663
https://tests.stockfishchess.org/tests/view/664ae88e830eb9f8866146f9

Second spsa with 30k / 120k games at 60+0.6:
https://tests.stockfishchess.org/tests/view/664be227830eb9f886615a36

Values found at 10k games at 60+0.6 also passed STC and LTC:
https://tests.stockfishchess.org/tests/view/664bf4bd830eb9f886615a72
https://tests.stockfishchess.org/tests/view/664c0905830eb9f886615abf

Passed STC:
https://tests.stockfishchess.org/tests/view/664c139e830eb9f886615af2
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 69408 W: 18216 L: 17842 D: 33350 Elo +1.87
Ptnml(0-2): 257, 8275, 17401, 8379, 392

Passed LTC:
https://tests.stockfishchess.org/tests/view/664cdaf7830eb9f886616a24
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 35466 W: 9075 L: 8758 D: 17633 Elo +3.11
Ptnml(0-2): 27, 3783, 9794, 4104, 25

closes https://github.com/official-stockfish/Stockfish/pull/5280

bench 1301287
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: MinetaS
Date: Tue May 21 08:46:12 2024 +0200
Timestamp: 1716273972

Remove cutoffCnt margin adjustment in razoring

Passed non-regression STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 65344 W: 16767 L: 16578 D: 31999 Elo +1.00
Ptnml(0-2): 198, 7557, 16987, 7718, 212
https://tests.stockfishchess.org/tests/view/664bd895830eb9f886615a26

Passed non-regression LTC:
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 35214 W: 8999 L: 8791 D: 17424 Elo +2.05
Ptnml(0-2): 16, 3804, 9760, 4010, 17
https://tests.stockfishchess.org/tests/view/664bead5830eb9f886615a52

closes https://github.com/official-stockfish/Stockfish/pull/5278

Bench: 1296223
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: FauziAkram
Date: Tue May 21 08:44:59 2024 +0200
Timestamp: 1716273899

Refine Evaluation Scaling with Piece-Specific Weights

Refine Evaluation Scaling with Piece-Specific Weights, instead of the simplified npm method.
I took the initial idea from Viren6 , as he worked on it in September of last year.
I worked on it, and tuned it, and now it passed both tests.

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 95712 W: 24731 L: 24325 D: 46656 Elo +1.47
Ptnml(0-2): 363, 11152, 24357, 11684, 300
https://tests.stockfishchess.org/tests/view/664b5493830eb9f886614af3

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 204480 W: 52167 L: 51501 D: 100812 Elo +1.13
Ptnml(0-2): 114, 22579, 56166, 23289, 92
https://tests.stockfishchess.org/tests/view/664b75dd830eb9f886614b44

closes https://github.com/official-stockfish/Stockfish/pull/5277

Bench: 1384337
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Shawn Xu
Date: Tue May 21 08:40:37 2024 +0200
Timestamp: 1716273637

Reduce When TTValue is Above Alpha

Passed STC:
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 53376 W: 13818 L: 13476 D: 26082 Elo +2.23
Ptnml(0-2): 156, 6212, 13626, 6522, 172
https://tests.stockfishchess.org/tests/view/664aa261830eb9f8866145e5

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 393444 W: 100096 L: 99042 D: 194306 Elo +0.93
Ptnml(0-2): 191, 43516, 108248, 44582, 185
https://tests.stockfishchess.org/tests/view/664ab54f830eb9f88661463c

closes https://github.com/official-stockfish/Stockfish/pull/5276

Bench: 1024562
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Michael Chaly
Date: Tue May 21 08:36:32 2024 +0200
Timestamp: 1716273392

Rescale pawn history updates

This patch is somewhat of a continuation of recent pawn history gainers.
It makes pawn history updates after search twice smaller. Since on average they make pawn history more negative offset is changed to lower value to remain average value approximately the same.

https://tests.stockfishchess.org/tests/view/664b3af9830eb9f886614aab
Passed STC:
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 170464 W: 44239 L: 43724 D: 82501 Elo +1.05
Ptnml(0-2): 523, 20278, 43128, 20767, 536

Passed LTC against pending PR :
https://tests.stockfishchess.org/tests/view/664b8c58830eb9f886614b64
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 98178 W: 25015 L: 24569 D: 48594 Elo +1.58
Ptnml(0-2): 48, 10769, 27005, 11223, 44

closes https://github.com/official-stockfish/Stockfish/pull/5275

Bench: 1343175
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: FauziAkram
Date: Tue May 21 08:27:16 2024 +0200
Timestamp: 1716272836

Use same shuffling Constant for both nets

Passed STC:
https://tests.stockfishchess.org/tests/view/664a42b15fc7b70b8817aeef
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 87840 W: 22759 L: 22594 D: 42487 Elo +0.65
Ptnml(0-2): 335, 10351, 22324, 10634, 276

Passed LTC:
https://tests.stockfishchess.org/tests/view/664a46995fc7b70b8817af02
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 163122 W: 41443 L: 41367 D: 80312 Elo +0.16
Ptnml(0-2): 105, 18154, 44927, 18310, 65

closes https://github.com/official-stockfish/Stockfish/pull/5273

bench: 1190174
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Michael Chaly
Date: Tue May 21 08:17:20 2024 +0200
Timestamp: 1716272240

Update correction history in case of successful null move pruning

Since null move pruning uses the same position it makes some sense to try to update correction history there in case of fail high.
Update value is 4 times less than normal update.

Passed STC:
https://tests.stockfishchess.org/tests/view/664a011cae57c1758ac5b4dd
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 419360 W: 108390 L: 107505 D: 203465 Elo +0.73
Ptnml(0-2): 1416, 49603, 106724, 50554, 1383

Passed LTC:
https://tests.stockfishchess.org/tests/view/664a53d95fc7b70b8817c65b
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 193518 W: 49076 L: 48434 D: 96008 Elo +1.15
Ptnml(0-2): 89, 21335, 53263, 21989, 83

closes https://github.com/official-stockfish/Stockfish/pull/5272

bench 1301487
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Tue May 21 08:13:25 2024 +0200
Timestamp: 1716272005

Re-eval only if smallnet output flips from simple eval

Recent attempts to change the smallnet nnue re-eval
threshold did not show much elo difference:
https://tests.stockfishchess.org/tests/view/664a29bb25a9058c4d21d53c
https://tests.stockfishchess.org/tests/view/664a299925a9058c4d21d53a

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/664a3ea95fc7b70b8817aee2
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 22304 W: 5905 L: 5664 D: 10735 Elo +3.75
Ptnml(0-2): 67, 2602, 5603, 2783, 97

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/664a43d35fc7b70b8817aef4
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 37536 W: 9667 L: 9460 D: 18409 Elo +1.92
Ptnml(0-2): 25, 4090, 10321, 4317, 15

closes https://github.com/official-stockfish/Stockfish/pull/5271

bench 1287409
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Stefan Geschwentner
Date: Tue May 21 08:11:17 2024 +0200
Timestamp: 1716271877

Simplify the recently introduced ply-based cmh bonus factor.

Replace it with a constant which is an approximation of the limit of the factor.

STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 120064 W: 30967 L: 30836 D: 58261 Elo +0.38
Ptnml(0-2): 421, 14238, 30608, 14319, 446
https://tests.stockfishchess.org/tests/view/6649d146b8fa20e74c39f4ad

LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 53856 W: 13719 L: 13530 D: 26607 Elo +1.22
Ptnml(0-2): 31, 5879, 14922, 6062, 34
https://tests.stockfishchess.org/tests/view/664a027fae57c1758ac5b4ee

closes https://github.com/official-stockfish/Stockfish/pull/5270

Bench: 1355618
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Michael Chaly
Date: Tue May 21 08:07:22 2024 +0200
Timestamp: 1716271642

Do more aggressive pawn history updates

Tweak of recent patch that made pawn history to update for move that caused a fail low - and setting up default value of it to -900. This patch makes it more aggressive - twice bigger updates and default value -1100.

Passed STC:
https://tests.stockfishchess.org/tests/view/6648c5d4308cceea45533b5d
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 235200 W: 61090 L: 60476 D: 113634 Elo +0.91
Ptnml(0-2): 763, 27952, 59651, 28376, 858

Passed LTC:
https://tests.stockfishchess.org/tests/view/664a1008ae57c1758ac5b523
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 20076 W: 5193 L: 4908 D: 9975 Elo +4.93
Ptnml(0-2): 7, 2105, 5534, 2380, 12

closes https://github.com/official-stockfish/Stockfish/pull/5268

Bench: 1590474
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Stefan Geschwentner
Date: Tue May 21 08:04:44 2024 +0200
Timestamp: 1716271484

Tweak continuation history bonus dependent on ply.

This patch is based on following tuning https://tests.stockfishchess.org/tests/view/6648b2eb308cceea45533abe by only using the tuned factors for the continuation history.

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 99904 W: 25865 L: 25457 D: 48582 Elo +1.42
Ptnml(0-2): 281, 11705, 25578, 12101, 287
https://tests.stockfishchess.org/tests/view/6648c136308cceea45533af8

Passed LTC:
LLR: 2.96 (-2.94,2.94) <0.50,2.50>
Total: 36402 W: 9362 L: 9039 D: 18001 Elo +3.08
Ptnml(0-2): 20, 3952, 9951, 4241, 37
https://tests.stockfishchess.org/tests/view/6648ee3cb8fa20e74c39f3fd

closes https://github.com/official-stockfish/Stockfish/pull/5267

Bench: 1917762
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: cj5716
Date: Tue May 21 07:58:16 2024 +0200
Timestamp: 1716271096

Simplify ClippedReLU

Removes some max calls

Some speedup stats, courtesy of @AndyGrant (albeit measured in an alternate implementation)
Dev 749240 nps
Base 748495 nps
Gain 0.100%
289936 games

STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 203040 W: 52213 L: 52179 D: 98648 Elo +0.06
Ptnml(0-2): 480, 20722, 59139, 20642, 537
https://tests.stockfishchess.org/tests/view/664805fe6dcff0d1d6b05f2c

closes #5261

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Viren6
Date: Sun May 19 09:37:22 2024 +0200
Timestamp: 1716104242

Revert "Simplify Away Quadruple Extensions"

This reverts commit 4edd1a3

The unusual result of (combined) +12.0 +- 3.7 in the 2 VVLTC simplification SPRTs ran was the result of base having only 64MB of hash instead of 512MB (Asymmetric hash).
Vizvezdenec was the one to notice this.

closes https://github.com/official-stockfish/Stockfish/pull/5265

bench 1404295

Co-Authored-By: Michael Chaly <26898827+>
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Muzhen Gaming
Date: Sat May 18 18:09:45 2024 +0200
Timestamp: 1716048585

VVLTC search tune

Tuned with 85k games at VVLTC.

VVLTC 1st sprt: https://tests.stockfishchess.org/tests/view/6648b836308cceea45533ad7
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 14880 W: 3890 L: 3652 D: 7338 Elo +5.56
Ptnml(0-2): 0, 1255, 4694, 1489, 2

VVLTC 2nd sprt: https://tests.stockfishchess.org/tests/view/6648c34f308cceea45533b4f
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 24984 W: 6502 L: 6235 D: 12247 Elo +3.71
Ptnml(0-2): 1, 2178, 7867, 2445, 1

closes https://github.com/official-stockfish/Stockfish/pull/5264

Bench: 1198142
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Sat May 18 18:08:39 2024 +0200
Timestamp: 1716048519

Simplify away adjustEval lambda

Now that only the shuffling constant differs between nets,
a lambda for adjusting eval is no longer needed.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/664806ca6dcff0d1d6b05f34
LLR: 2.99 (-2.94,2.94) <-1.75,0.25>
Total: 31552 W: 8175 L: 7959 D: 15418 Elo +2.38
Ptnml(0-2): 76, 3180, 9065, 3362, 93

closes https://github.com/official-stockfish/Stockfish/pull/5260

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Shawn Xu
Date: Sat May 18 09:32:21 2024 +0200
Timestamp: 1716017541

Simplify Away Quadruple Extensions

serendipitous gainer

Passed LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 95472 W: 24176 L: 24031 D: 47265 Elo +0.53
Ptnml(0-2): 52, 10533, 26414, 10692, 45
https://tests.stockfishchess.org/tests/live_elo/6647fa596dcff0d1d6b05efa

Passed VVLTC 70+7 th 7:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 6772 W: 1793 L: 1583 D: 3396 Elo +10.78
Ptnml(0-2): 0, 502, 2172, 712, 0
https://tests.stockfishchess.org/tests/live_elo/6648277a6dcff0d1d6b05ffb

Passed VVLTC 70+7 th 7 (2x):
https://tests.stockfishchess.org/tests/view/66484c896dcff0d1d6b0619d
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 5424 W: 1469 L: 1254 D: 2701 Elo +13.78
Ptnml(0-2): 0, 394, 1710, 607, 1

closes https://github.com/official-stockfish/Stockfish/pull/5259

Bench: 1441794
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Sat May 18 09:29:26 2024 +0200
Timestamp: 1716017366

Use one nnue pawn count multiplier

Switch to the value used by the main net.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/6647e8096dcff0d1d6b05e96
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 51040 W: 13249 L: 13044 D: 24747 Elo +1.40
Ptnml(0-2): 139, 6029, 13016, 6160, 176

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/6647f4a46dcff0d1d6b05eea
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 20460 W: 5195 L: 4972 D: 10293 Elo +3.79
Ptnml(0-2): 8, 2178, 5637, 2397, 10

https://github.com/official-stockfish/Stockfish/pull/5258

bench 1887462
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Shawn Xu
Date: Sat May 18 09:26:42 2024 +0200
Timestamp: 1716017202

Tweak NMP Formula

Passed STC:
LLR: 2.99 (-2.94,2.94) <0.00,2.00>
Total: 241728 W: 62440 L: 61811 D: 117477 Elo +0.90
Ptnml(0-2): 914, 28467, 61458, 29126, 899
https://tests.stockfishchess.org/tests/live_elo/6645992993ce6da3e93b5b99

Passed LTC:
LLR: 2.96 (-2.94,2.94) <0.50,2.50>
Total: 167850 W: 42620 L: 42030 D: 83200 Elo +1.22
Ptnml(0-2): 82, 18412, 46354, 18988, 89
https://tests.stockfishchess.org/tests/live_elo/6647c5726dcff0d1d6b05dd3

closes https://github.com/official-stockfish/Stockfish/pull/5257

Bench: 1636018
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: FauziAkram
Date: Sat May 18 09:24:23 2024 +0200
Timestamp: 1716017063

Early Exit in Bitboards::sliding_attack()

he original code checks for occupancy within the loop condition. By moving this check inside the loop and adding an early exit condition, we can avoid unnecessary iterations if a blocking piece is encountered.

Passed stc:
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 127200 W: 33129 L: 32700 D: 61371 Elo +1.17
Ptnml(0-2): 424, 13243, 35826, 13694, 413
https://tests.stockfishchess.org/tests/view/664646006dcff0d1d6b05bca

closes https://github.com/official-stockfish/Stockfish/pull/5256

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Sat May 18 09:21:00 2024 +0200
Timestamp: 1716016860

Move smallnet threshold logic into a function

Now that the smallnet threshold is no longer a constant,
use a function to organize it with other eval code.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/66459fa093ce6da3e93b5ba2
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 217600 W: 56281 L: 56260 D: 105059 Elo +0.03
Ptnml(0-2): 756, 23787, 59729, 23736, 792

closes https://github.com/official-stockfish/Stockfish/pull/5255

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Sat May 18 09:19:10 2024 +0200
Timestamp: 1716016750

Update default main net to nn-c721dfca8cd3.nnue

Created by first retraining the spsa-tuned main net `nn-ae6a388e4a1a.nnue` with:
- using v6-dd data without bestmove captures removed
- addition of T80 mar2024 data
- increasing loss by 20% when Q is too high
- torch.compile changes for marginal training speed gains

And then SPSA tuning weights of epoch 899 following methods described in:
https://github.com/official-stockfish/Stockfish/pull/5149

This net was reached at 92k out of 120k steps in this 70+0.7 th 7 SPSA tuning run:
https://tests.stockfishchess.org/tests/view/66413b7df9f4e8fc783c9bbb
Thanks to @Viren6 for suggesting usage of:
- c value 4 for the weights
- c value 128 for the biases

Scripts for automating applying fishtest spsa params to exporting tuned .nnue are in:
https://github.com/linrock/nnue-tools/tree/master/spsa

Before spsa tuning, epoch 899 was nn-f85738aefa84.nnue
https://tests.stockfishchess.org/tests/view/663e5c893a2f9702074bc167

After initially training with max-epoch 800, training was resumed with max-epoch 1000.

```
experiment-name: 3072--S11--more-data-v6-dd-t80-mar2024--see-ge0-20p-more-loss-high-q-sk28-l8
nnue-pytorch-branch: linrock/nnue-pytorch/3072-r21-skip-more-wdl-see-ge0-20p-more-loss-high-q-torch-compile-more

start-from-engine-test-net: False
start-from-model: /data/config/apr2024-3072/nn-ae6a388e4a1a.nnue

early-fen-skipping: 28
training-dataset:
/data/S11-mar2024/:
- leela96.v2.min.binpack

- test60-2021-11-12-novdec-12tb7p.v6-dd.min.binpack
- test78-2022-01-to-05-jantomay-16tb7p.v6-dd.min.binpack

- test80-2022-06-jun-16tb7p.v6-dd.min.binpack

- test80-2022-08-aug-16tb7p.v6-dd.min.binpack
- test80-2022-09-sep-16tb7p.v6-dd.min.binpack

- test80-2023-01-jan-16tb7p.v6-sk20.min.binpack
- test80-2023-02-feb-16tb7p.v6-sk20.min.binpack
- test80-2023-03-mar-2tb7p.v6-sk16.min.binpack
- test80-2023-04-apr-2tb7p.v6-sk16.min.binpack
- test80-2023-05-may-2tb7p.v6.min.binpack

# https://github.com/official-stockfish/Stockfish/pull/4782
- test80-2023-06-jun-2tb7p.binpack
- test80-2023-07-jul-2tb7p.binpack

# https://github.com/official-stockfish/Stockfish/pull/4972
- test80-2023-08-aug-2tb7p.v6.min.binpack
- test80-2023-09-sep-2tb7p.binpack
- test80-2023-10-oct-2tb7p.binpack

# S9 new data: https://github.com/official-stockfish/Stockfish/pull/5056
- test80-2023-11-nov-2tb7p.binpack
- test80-2023-12-dec-2tb7p.binpack

# S10 new data: https://github.com/official-stockfish/Stockfish/pull/5149
- test80-2024-01-jan-2tb7p.binpack
- test80-2024-02-feb-2tb7p.binpack

# S11 new data
- test80-2024-03-mar-2tb7p.binpack

/data/filt-v6-dd/:
- test77-dec2021-16tb7p-filter-v6-dd.binpack
- test78-juntosep2022-16tb7p-filter-v6-dd.binpack
- test79-apr2022-16tb7p-filter-v6-dd.binpack
- test79-may2022-16tb7p-filter-v6-dd.binpack
- test80-jul2022-16tb7p-filter-v6-dd.binpack
- test80-oct2022-16tb7p-filter-v6-dd.binpack
- test80-nov2022-16tb7p-filter-v6-dd.binpack

num-epochs: 1000

lr: 4.375e-4
gamma: 0.995
start-lambda: 0.8
end-lambda: 0.7
```

Training data can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch899.nnue : 4.6 +/- 1.4

Passed STC:
https://tests.stockfishchess.org/tests/view/6645454893ce6da3e93b31ae
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 95232 W: 24598 L: 24194 D: 46440 Elo +1.47
Ptnml(0-2): 294, 11215, 24180, 11647, 280

Passed LTC:
https://tests.stockfishchess.org/tests/view/6645522d93ce6da3e93b31df
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 320544 W: 81432 L: 80524 D: 158588 Elo +0.98
Ptnml(0-2): 164, 35659, 87696, 36611, 142

closes https://github.com/official-stockfish/Stockfish/pull/5254

bench 1995552
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Rak Laptudirm
Date: Sat May 18 09:18:14 2024 +0200
Timestamp: 1716016694

Improve comment

closes https://github.com/official-stockfish/Stockfish/pull/5249

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Thu May 16 14:19:28 2024 +0200
Timestamp: 1715861968

Lower smallnet threshold linearly as pawn count decreases

Passed STC:
https://tests.stockfishchess.org/tests/view/6644f677324e96f42f89d894
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 377920 W: 97135 L: 96322 D: 184463 Elo +0.75
Ptnml(0-2): 1044, 44259, 97588, 44978, 1091

Passed LTC:
https://tests.stockfishchess.org/tests/view/664548af93ce6da3e93b31b3
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 169056 W: 42901 L: 42312 D: 83843 Elo +1.21
Ptnml(0-2): 58, 18538, 46753, 19115, 64

closes https://github.com/official-stockfish/Stockfish/pull/5252

Bench: 1991750
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Muzhen Gaming
Date: Thu May 16 14:17:53 2024 +0200
Timestamp: 1715861873

Revert "Reduce more when improving and ttvalue is lower than alpha"

The patch regressed significantly at longer time controls.

Passed VLTC:
https://tests.stockfishchess.org/tests/view/6644c7a2bc537f5619453096
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 43336 W: 11177 L: 10884 D: 21275 Elo +2.35
Ptnml(0-2): 3, 4432, 12507, 4721, 5

Passed VVLTC:
https://tests.stockfishchess.org/tests/view/66450c974aa4fa9a83b6d0b0
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 32394 W: 8350 L: 8072 D: 15972 Elo +2.98
Ptnml(0-2): 2, 2798, 10317, 3080, 0

closes https://github.com/official-stockfish/Stockfish/pull/5251

Bench: 1594188
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Linmiao Xu
Date: Thu May 16 14:16:54 2024 +0200
Timestamp: 1715861814

Use same nnue divisor for both nets

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/6643ceeabc537f56194506f6
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 224800 W: 57910 L: 57896 D: 108994 Elo +0.02
Ptnml(0-2): 673, 26790, 57519, 26686, 732

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/6643ff15bc537f5619451719
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 347658 W: 87574 L: 87688 D: 172396 Elo -0.11
Ptnml(0-2): 207, 39004, 95488, 38956, 174

closes https://github.com/official-stockfish/Stockfish/pull/5250

Bench: 1804704
see source

next page >