Fabian Fichter [Sun, 28 Nov 2021 19:31:31 +0000]
Fix disambiguation of special moves
Closes #408.
Fabian Fichter [Sun, 28 Nov 2021 13:30:52 +0000]
Support time unit switch for UCCI
Closes #406.
Fabian Fichter [Sat, 13 Nov 2021 15:26:29 +0000]
Add unit test to cover bugfix
Fabian Fichter [Sat, 13 Nov 2021 12:15:07 +0000]
Fix castling rights for passing moves
Snowmoondaphne [Wed, 10 Nov 2021 01:34:14 +0000]
Update variants.ini
I'm really sorry to tell you this,,,
The positions of Black's Queen and Cardinal have been swapped in Pandemonium. Therefore, the definition has changed and it is necessary to modify it
[pandemonium]
variantTemplate = shogi
pieceToCharTable = PNBRFSA.UV.++++++++.++Kpnbrfsa.uv.++++++++.++k
maxFile = 9
maxRank = 9
pocketSize = 9
startFen = rnbsksbnr/2+f1+u1+a2/p1p1p1p1p/4v4/9/4V4/P1P1P1P1P/2+F1+U1+A2/RNBSKSBNR[] w - - 0 1
customPiece1 = o:NA
customPiece2 = s:WF
customPiece3 = u:D
customPiece4 = w:DWF
cast = false
pieceDrops = true
capturesToHand = true
immobilityIllegal = true
soldier = p
knight = n
bishop = b
rook = r
king = k
queen = q
commoner = g
dragonHorse = h
bers = d
alfil = a
archbishop = c
chancellor = m
fers = f
wazir = v
centaur = t
promotionRank = 7
promotedPieceType = p:g n:o b:h r:d a:c v:m f:q s:w u:t
doubleStep = false
perpetualCheckIllegal = true
nMoveRule = 0
nFoldValue = loss
stalemateValue = loss
Could you please modify the definition like this?
Sorry again for the troublesome request,,,
Fabian Fichter [Fri, 5 Nov 2021 23:05:38 +0000]
Remove redundancy from S-Chess FEN
Closes #83.
Snowmoondaphne [Tue, 26 Oct 2021 10:02:40 +0000]
Update variants.ini
Add a variant
Fabian Fichter [Sat, 6 Nov 2021 16:59:06 +0000]
Simplify configuration of custom royal pieces
Use the same syntax for custom king as for other custom pieces.
Closes #363.
Fabian Fichter [Sat, 6 Nov 2021 10:14:01 +0000]
Use tandem pawn notation for double tandems
In WXF notation +/- disambiguation can be ambiguous
when there is more than one tandem of pieces/pawns.
Closes #395.
Fabian Fichter [Sat, 6 Nov 2021 14:21:38 +0000]
Support validation of lichess 3check FENs
Closes #288.
Fabian Fichter [Sat, 6 Nov 2021 12:59:42 +0000]
Validate gating information in FEN (#288)
Fabian Fichter [Fri, 5 Nov 2021 21:25:39 +0000]
Validate variants.ini syntax
Closes #388.
SriMethan [Wed, 3 Nov 2021 13:06:23 +0000]
Mention lishogi fairy stockfish bot in README.md
Fabian Fichter [Mon, 1 Nov 2021 17:44:08 +0000]
Add optimized dev builds
Fabian Fichter [Thu, 28 Oct 2021 17:10:34 +0000]
Remove check indicator from WXF notation
Fabian Fichter [Fri, 15 Oct 2021 16:42:27 +0000]
Disable NNUE in case of virtual pieces
Fabian Fichter [Fri, 15 Oct 2021 11:14:16 +0000]
Adjust max NNUE pieces for two board variants
Fabian Fichter [Thu, 14 Oct 2021 06:55:00 +0000]
Update betza notation in docs
Fabian Fichter [Thu, 14 Oct 2021 06:41:17 +0000]
Update authors list
Fabian Fichter [Sat, 9 Oct 2021 16:08:28 +0000]
Extend perft test coverage
Michael Taktikos [Sat, 9 Oct 2021 13:52:27 +0000]
Fixed Betza notation of Breakthrough piece
Avoid illegal moves (captures forward) in Breakthrough
Terry Hearst [Fri, 8 Oct 2021 13:16:39 +0000]
Expose additional game-end functions in ffish.js (#362)
Ada Joule [Sun, 3 Oct 2021 05:11:58 +0000]
Expose position.fen's showPromoted and countStarted in ffish.js
Fabian Fichter [Sat, 2 Oct 2021 21:46:35 +0000]
Handle placement corner cases more naturally
No functional change for legal positions,
just generalizes more naturally to artifical positions.
gbtami [Sat, 2 Oct 2021 15:00:04 +0000]
Bump pyffish version
Ada Joule [Fri, 1 Oct 2021 18:32:19 +0000]
Correct ASEAN's counting rules
- Khon and Knight don't need a Met to start the counting
- Counting only starts after all pawns are promoted
Fabian Fichter [Fri, 1 Oct 2021 10:01:02 +0000]
Support NNUE with varying king dimensions
This adds dedicated NNUE support for variants where kings
only have access to a limited set of squares, like Xiangqi,
or are missing entirely, like in antichess.
Closes #346.
Fabian Fichter [Mon, 27 Sep 2021 21:15:48 +0000]
Support spaces in variants.ini path
* Allow whitespace characters in `load` command.
* Support loading of multiple files separated by `:`/`;`.
Closes #361.
Fabian Fichter [Mon, 27 Sep 2021 19:59:18 +0000]
Bump pyffish version
Fabian Fichter [Mon, 27 Sep 2021 19:41:51 +0000]
Limit pawn moves to variant board size
Closes #372.
Fabian Fichter [Sun, 26 Sep 2021 15:54:14 +0000]
More robust shogi FEN parsing
Closes #370.
Terry Hearst [Fri, 17 Sep 2021 07:03:25 +0000]
Autoload variants.ini if specified in by an environment variable (#360)
CouchTomato87 [Wed, 15 Sep 2021 13:35:58 +0000]
Corrected Tori n-fold repetition. (#355)
Fabian Fichter [Sun, 12 Sep 2021 15:04:01 +0000]
Bump pyffish version
Ada Joule [Sun, 12 Sep 2021 12:15:15 +0000]
Ignore countStarted when either side has only one piece
Fabian Fichter [Fri, 10 Sep 2021 18:59:57 +0000]
Restore development version
Fabian Fichter [Fri, 10 Sep 2021 14:49:27 +0000]
Fairy-Stockfish 14
bench: 5903196
Fabian Fichter [Tue, 7 Sep 2021 17:09:18 +0000]
Fix insufficient material for atomic
Closes #352.
Fabian Fichter [Sat, 4 Sep 2021 21:08:24 +0000]
Fix incremental update for hand pieces
Fabian Fichter [Sat, 4 Sep 2021 16:13:44 +0000]
Bump pyffish version
Fabian Fichter [Sat, 4 Sep 2021 14:45:43 +0000]
Fix pawn evasions for cannon checks
Fabian Fichter [Sat, 4 Sep 2021 07:53:35 +0000]
Support NNUE for placement variants
Use classical evaluation during setup phase to enable NNUE usage.
Fabian Fichter [Fri, 3 Sep 2021 12:08:31 +0000]
Update readme
Fabian Fichter [Fri, 3 Sep 2021 10:16:22 +0000]
Enable NNUE pockets for gating variants
Fabian Fichter [Fri, 3 Sep 2021 07:53:04 +0000]
Simplify addition of variants
Bajusz Tamás [Wed, 1 Sep 2021 09:02:28 +0000]
Bump pyffish version (#350)
Fabian Fichter [Sun, 29 Aug 2021 09:16:53 +0000]
Generalize incremental computation for hand pieces
This allows to apply incremental NNUE computation not only to drop variants,
but also to S-Chess gating and drop variants with blast captures.
Fabian Fichter [Sat, 28 Aug 2021 12:04:44 +0000]
Support re-usage of NNUE nets for other variants
Define aliases for compatible variants, so that
they can re-use existing NNUE files without requiring
to create symlinks or copies of the evaluation files.
Fabian Fichter [Sat, 28 Aug 2021 08:42:16 +0000]
Support pieces in hand for NNUE
This enables to consider pieces in hand in NNUE evaluation.
Fabian Fichter [Fri, 30 Jul 2021 13:18:37 +0000]
Generalize NNUE architecture
Remove restrictions on board size and piece types
in order to enable NNUE for most supported variants.
This change modifies the indexing for fairy pieces,
therefore existing networks for fairy variants, currently only Makruk,
will not work properly any more and need to be replaced.
Limitations that are not addressed as part of this are:
- A "king" piece type with a constant count of 1 is required.
- Pockets are not considered.
Major Mink [Fri, 6 Aug 2021 12:08:03 +0000]
Fixed blasted pieces incorrectly updating key's castling rights
Fabian Fichter [Thu, 5 Aug 2021 14:42:54 +0000]
Prevent division by zero in extinction bonus
Fixes #340.
Bajusz Tamás [Sat, 31 Jul 2021 09:06:39 +0000]
Fix Orda Mirror promotionPieceTypes definition (#339)
Bajusz Tamás [Fri, 30 Jul 2021 15:46:33 +0000]
Add Python3.9 support to appveyor (#338)
Fabian Fichter [Thu, 29 Jul 2021 21:19:46 +0000]
Fix insufficient material handling for makpong
Closes #335.
QueensGambit [Thu, 29 Jul 2021 17:44:47 +0000]
Update ffish.js version
Bajusz Tamás [Thu, 29 Jul 2021 07:09:25 +0000]
Bump pyffish version (#334)
Fabian Fichter [Tue, 27 Jul 2021 20:15:31 +0000]
Fix variant initialization for libraries
Fabian Fichter [Sun, 25 Jul 2021 19:06:52 +0000]
Fix NNUE evaluation trace for variants
Fabian Fichter [Tue, 20 Jul 2021 18:05:05 +0000]
Fix memory leak of CECP state
Closes #331.
Fabian Fichter [Tue, 20 Jul 2021 17:16:35 +0000]
Reintroduce tempo for drop variants
crazyhouse STC
LLR: 2.97 (-2.94,2.94) [0.00,10.00]
Total: 1439 W: 752 L: 632 D: 55
http://www.variantfishtest.org:6543/tests/view/
60f4410abf545bb49fff86bc
crazyhouse LTC
LLR: 2.96 (-2.94,2.94) [0.00,10.00]
Total: 1403 W: 741 L: 621 D: 41
http://www.variantfishtest.org:6543/tests/view/
60f51db7bf545bb49fff86c6
Closes #332.
Fabian Fichter [Sun, 18 Jul 2021 14:23:30 +0000]
Merge official-stockfish/master
bench: 5903196
Fabian Fichter [Sun, 18 Jul 2021 13:56:13 +0000]
Merge official-stockfish/master
bench: 6603046
Fabian Fichter [Sun, 18 Jul 2021 13:44:41 +0000]
Merge official-stockfish/master
bench: 6215456
Fabian Fichter [Sun, 18 Jul 2021 13:37:21 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Sun, 18 Jul 2021 13:30:52 +0000]
Merge official-stockfish/master
bench: 6141390
Fabian Fichter [Sun, 18 Jul 2021 13:26:37 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Sun, 18 Jul 2021 11:56:42 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Sun, 18 Jul 2021 11:51:44 +0000]
Merge official-stockfish/master
bench: 4890740
Fabian Fichter [Sat, 17 Jul 2021 20:18:28 +0000]
Merge official-stockfish/master
bench: 4456706
Fabian Fichter [Sat, 17 Jul 2021 20:13:34 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Sat, 17 Jul 2021 20:01:22 +0000]
Merge official-stockfish/master
bench: 4590190
Fabian Fichter [Sat, 17 Jul 2021 19:44:32 +0000]
Merge official-stockfish/master
bench: 4858668
Fabian Fichter [Sat, 17 Jul 2021 19:37:39 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Sat, 17 Jul 2021 19:31:06 +0000]
Merge official-stockfish/master
bench: 4838140
Fabian Fichter [Fri, 16 Jul 2021 17:19:34 +0000]
Fix NNUE index for commoners
Fabian Fichter [Fri, 16 Jul 2021 10:11:09 +0000]
Merge official-stockfish/master
bench: 4850077
Joost VandeVondele [Sat, 3 Jul 2021 07:20:06 +0000]
Restore development version
No functional change
Joost VandeVondele [Mon, 28 Jun 2021 19:46:04 +0000]
Stockfish 14
Official release version of Stockfish 14
Bench: 4770936
---
Today, we have the pleasure to announce Stockfish 14.
As usual, downloads will be freely available at https://stockfishchess.org
The engine is now significantly stronger than just a few months ago,
and wins four times more game pairs than it loses against the previous
release version [0]. Stockfish 14 is now at least 400 Elo ahead of
Stockfish 7, a top engine in 2016 [1]. During the last five years,
Stockfish has thus gained about 80 Elo per year.
Stockfish 14 evaluates positions more accurately than Stockfish 13 as
a result of two major steps forward in defining and training the
efficiently updatable neural network (NNUE) that provides the evaluation
for positions.
First, the collaboration with the Leela Chess Zero team - announced
previously [2] - has come to fruition. The LCZero team has provided a
collection of billions of positions evaluated by Leela that we have
combined with billions of positions evaluated by Stockfish to train the
NNUE net that powers Stockfish 14. The fact that we could use and combine
these datasets freely was essential for the progress made and demonstrates
the power of open source and open data [3].
Second, the architecture of the NNUE network was significantly updated:
the new network is not only larger, but more importantly, it deals better
with large material imbalances and can specialize for multiple phases of
the game [4]. A new project, kick-started by Gary Linscott and
Tomasz Sobczyk, led to a GPU accelerated net trainer written in
pytorch.[5] This tool allows for training high-quality nets in a couple
of hours.
Finally, this release features some search refinements, minor bug
fixes and additional improvements. For example, Stockfish is now about
90 Elo stronger for chess960 (Fischer random chess) at short time control.
The Stockfish project builds on a thriving community of enthusiasts
(thanks everybody!) that contribute their expertise, time, and resources
to build a free and open-source chess engine that is robust, widely
available, and very strong. We invite our chess fans to join the fishtest
testing framework and programmers to contribute to the project on
github [6].
Stay safe and enjoy chess!
The Stockfish team
[0] https://tests.stockfishchess.org/tests/view/
60dae5363beab81350aca077
[1] https://nextchessmove.com/dev-builds
[2] https://stockfishchess.org/blog/2021/stockfish-13/
[3] https://lczero.org/blog/2021/06/the-importance-of-open-data/
[4] https://github.com/official-stockfish/Stockfish/commit/
e8d64af1
[5] https://github.com/glinscott/nnue-pytorch/
[6] https://stockfishchess.org/get-involved/
Fabian Fichter [Fri, 2 Jul 2021 08:06:24 +0000]
Update js makefile
Fabian Fichter [Thu, 1 Jul 2021 22:28:57 +0000]
Merge official-stockfish/master
bench: 4835554
Fabian Fichter [Thu, 1 Jul 2021 22:20:42 +0000]
Merge official-stockfish/master
Fabian Fichter [Thu, 1 Jul 2021 22:11:00 +0000]
Merge official-stockfish/master
bench: 4806948
Fabian Fichter [Thu, 1 Jul 2021 21:58:55 +0000]
Merge official-stockfish/master
No functional change.
Fabian Fichter [Thu, 1 Jul 2021 21:42:05 +0000]
Merge official-stockfish/master
bench: 5885720
Brad Knox [Tue, 29 Jun 2021 06:40:16 +0000]
Update Top CPU Contributors
closes https://github.com/official-stockfish/Stockfish/pull/3595
No functional change
SFisGOD [Mon, 28 Jun 2021 06:58:51 +0000]
Update default net to nn-
3475407dc199.nnue
Optimization of eight subnetwork output layers of Michael's nn-
190f102a22c3.nnue using SPSA
https://tests.stockfishchess.org/tests/view/
60d5510642a522cc50282ef3
Parameters: A total of 256 net weights and 8 net biases were tuned
New best values: The raw values at the end of the tuning run were used (800k games, 5 seconds TC)
Settings: default ck value and SPSA A is 30,000 (3.75% of the total number of games)
STC:
LLR: 2.94 (-2.94,2.94) <-0.50,2.50>
Total: 29064 W: 2435 L: 2269 D: 24360
Ptnml(0-2): 72, 1857, 10505, 2029, 69
https://tests.stockfishchess.org/tests/view/
60d8ea123beab81350ac9eb6
LTC:
LLR: 2.93 (-2.94,2.94) <0.50,3.50>
Total: 61848 W: 2055 L: 1884 D: 57909
Ptnml(0-2): 18, 1708, 27310, 1861, 27
https://tests.stockfishchess.org/tests/view/
60d8f0393beab81350ac9ec6
closes https://github.com/official-stockfish/Stockfish/pull/3593
Bench: 4770936
MichaelB7 [Sun, 27 Jun 2021 15:26:09 +0000]
Make net nn-
956480d8378f.nnue the default
Trained with the pytorch trainer: https://github.com/glinscott/nnue-pytorch
python train.py i:/bin/all.binpack i:/bin/all.binpack --gpus 1 --threads 4 --num-workers 30 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 3 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed %random%%random% --default_root_dir exp/run_18 --resume-from-model ./pt/nn-
75980ca503c6.pt
This run is thus started from a previous master net.
all.binpack equaled 4 parts Wrong_NNUE_2.binpack https://drive.google.com/file/d/1seGNOqcVdvK_vPNq98j-zV3XPE5zWAeq/view?usp=sharing plus two parts of Training_Data.binpack https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing
Each set was concatenated together - making one large Wrong_NNUE 2 binpack and one large Training so the were approximately equal in size. They were then interleaved together. The idea was to give Wrong_NNUE.binpack closer to equal weighting with the Training_Data binpack
passed STC:
https://tests.stockfishchess.org/tests/view/
60d0c0a7a8ec07dc34c072b2
LLR: 2.93 (-2.94,2.94) <-0.50,2.50>
Total: 18440 W: 1693 L: 1531 D: 15216
Ptnml(0-2): 67, 1225, 6464, 1407, 57
passed LTC:
https://tests.stockfishchess.org/tests/view/
60d762793beab81350ac9d72
LLR: 2.98 (-2.94,2.94) <0.50,3.50>
Total: 93120 W: 3152 L: 2933 D: 87035
Ptnml(0-2): 48, 2581, 41076, 2814, 41
passed LTC (rebased branch to current master):
https://tests.stockfishchess.org/tests/view/
60d85eeb3beab81350ac9e2b
LLR: 2.96 (-2.94,2.94) <0.50,3.50>
Total: 42688 W: 1347 L: 1206 D: 40135
Ptnml(0-2): 14, 1097, 18981, 1238, 14.
closes https://github.com/official-stockfish/Stockfish/pull/3592
Bench: 4906727
Joost VandeVondele [Wed, 23 Jun 2021 05:23:21 +0000]
Update WDL model for NNUE
This updates the WDL model based on the LTC statistics in June this year (10M games),
so from pre-NNUE to NNUE based results.
(for old results see, https://github.com/official-stockfish/Stockfish/pull/2778)
As before the fit by the model to the data is quite good.
closes https://github.com/official-stockfish/Stockfish/pull/3582
No functional change
bmc4 [Tue, 22 Jun 2021 22:33:14 +0000]
Simplify Reductions Initialization
passed
STC:
LLR: 2.94 (-2.94,2.94) <-2.50,0.50>
Total: 45032 W: 3600 L: 3518 D: 37914
Ptnml(0-2): 111, 2893, 16435, 2957, 120
https://tests.stockfishchess.org/tests/view/
60d2655d40925195e7a6c527
LTC:
LLR: 3.00 (-2.94,2.94) <-2.50,0.50>
Total: 25728 W: 786 L: 722 D: 24220
Ptnml(0-2): 5, 650, 11494, 706, 9
https://tests.stockfishchess.org/tests/view/
60d2b14240925195e7a6c577
closes https://github.com/official-stockfish/Stockfish/pull/3584
bench: 4602977
Fabian Fichter [Sat, 26 Jun 2021 20:44:43 +0000]
Add missing include
Fabian Fichter [Sat, 26 Jun 2021 18:26:01 +0000]
Merge official-stockfish/master
Fabian Fichter [Sat, 26 Jun 2021 14:04:59 +0000]
Merge official-stockfish/master
bench: 5173306
Stéphane Nicolet [Tue, 22 Jun 2021 07:08:37 +0000]
Detect fortresses a little bit quicker
In the so-called "hybrid" method of evaluation of current master, we use the
classical eval (because of its speed) instead of the NNUE eval when the classical
material balance approximation hints that the position is "winning enough" to
rely on the classical eval.
This trade-off idea between speed and accuracy works well in general, but in
some fortress positions the classical eval is just bad. So in shuffling branches
of the search tree, we (slowly) increase the thresehold so that eventually we
don't trust classical anymore and switch to NNUE evaluation.
This patch increases that threshold faster, so that we switch to NNUE quicker
in shuffling branches. Idea is to incite Stockfish to spend less time in fortresses
lines in the search tree, and spend more time searching the critical lines.
passed STC:
LLR: 2.96 (-2.94,2.94) <-0.50,2.50>
Total: 47872 W: 3908 L: 3720 D: 40244
Ptnml(0-2): 122, 3053, 17419, 3199, 143
https://tests.stockfishchess.org/tests/view/
60cef34b457376eb8bcab79d
passed LTC:
LLR: 2.93 (-2.94,2.94) <0.50,3.50>
Total: 73616 W: 2326 L: 2143 D: 69147
Ptnml(0-2): 21, 1940, 32705, 2119, 23
https://tests.stockfishchess.org/tests/view/
60cf6d842114332881e73528
Retested at LTC against lastest master:
LLR: 2.93 (-2.94,2.94) <0.50,3.50>
Total: 18264 W: 642 L: 532 D: 17090
Ptnml(0-2): 6, 479, 8055, 583, 9
https://tests.stockfishchess.org/tests/view/
60d18cd540925195e7a6c351
closes https://github.com/official-stockfish/Stockfish/pull/3578
Bench: 5139233
MichaelB7 [Mon, 21 Jun 2021 12:10:35 +0000]
Make net nn-
190f102a22c3.nnue the default net.
Trained with the pytorch trainer: https://github.com/glinscott/nnue-pytorch
python train.py i:/bin/all.binpack i:/bin/all.binpack --gpus 1 --threads 4 --num-workers 30 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 3 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed %random%%random% --default_root_dir exp/run_17 --resume-from-model ./pt/nn-
75980ca503c6.pt
This run is thus started from the previous master net.
all.binpack equaled 4 parts Wrong_NNUE_2.binpack https://drive.google.com/file/d/1seGNOqcVdvK_vPNq98j-zV3XPE5zWAeq/view?usp=sharing plus two parts of Training_Data.binpack https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing
Each set was concatenated together - making one large Wrong_NNUE 2 binpack and one large Training so the were approximately equal in size. They were then interleaved together. The idea was to give Wrong_NNUE.binpack closer to equal weighting with the Training_Data binpack
passed LTC
https://tests.stockfishchess.org/tests/view/
60d09f52b4c17000d679517f
LLR: 2.93 (-2.94,2.94) <0.50,3.50>
Total: 32184 W: 1100 L: 970 D: 30114
Ptnml(0-2): 10, 878, 14193, 994, 17
passed STC
https://tests.stockfishchess.org/tests/view/
60d086c02114332881e7368e
LLR: 2.93 (-2.94,2.94) <-0.50,2.50>
Total: 11360 W: 1056 L: 906 D: 9398
Ptnml(0-2): 25, 735, 4026, 853, 41
closes https://github.com/official-stockfish/Stockfish/pull/3576
Bench: 4631244
Joost VandeVondele [Mon, 21 Jun 2021 06:23:50 +0000]
Fix build error on OSX
directly use integer version for cp calculation.
fixes https://github.com/official-stockfish/Stockfish/issues/3573
closes https://github.com/official-stockfish/Stockfish/pull/3574
No functional change
Stéphane Nicolet [Wed, 16 Jun 2021 05:23:26 +0000]
Remove the Contempt UCI option
This patch removes the UCI option for setting Contempt in classical evaluation.
It is exactly equivalent to using Contempt=0 for the UCI contempt value and keeping
the dynamic part in the algo (renaming this dynamic part `trend` to better describe
what it does). We have tried quite hard to implement a working Contempt feature for
NNUE but nothing really worked, so it is probably time to give up.
Interested chess fans wishing to keep playing with the UCI option for Contempt and
use it with the classical eval are urged to download the version tagged "SF_Classical"
of Stockfish (dated 31 July 2020), as it was the last version where our search
algorithm was tuned for the classical eval and is probably our strongest classical
player ever: https://github.com/official-stockfish/Stockfish/tags
Passed STC:
LLR: 2.95 (-2.94,2.94) <-2.50,0.50>
Total: 72904 W: 6228 L: 6175 D: 60501
Ptnml(0-2): 221, 5006, 25971, 5007, 247
https://tests.stockfishchess.org/tests/view/
60c98bf9457376eb8bcab18d
Passed LTC:
LLR: 2.93 (-2.94,2.94) <-2.50,0.50>
Total: 45168 W: 1601 L: 1547 D: 42020
Ptnml(0-2): 38, 1331, 19786, 1397, 32
https://tests.stockfishchess.org/tests/view/
60c9c7fa457376eb8bcab1bb
closes https://github.com/official-stockfish/Stockfish/pull/3575
Bench: 4947716
Stéphane Nicolet [Sun, 20 Jun 2021 08:29:20 +0000]
Keep more pawns and pieces when attacking
This patch increase the weight of pawns and pieces from 28 to 32
in the scaling formula we apply to the output of the NNUE pure eval.
Increasing this gradient for pawns and pieces means that Stockfish
will try a little harder to keep material when she has the advantage,
and try a little bit harder to escape into an endgame when she is
under pressure.
STC:
LLR: 2.93 (-2.94,2.94) <-0.50,2.50>
Total: 53168 W: 4371 L: 4177 D: 44620
Ptnml(0-2): 160, 3389, 19283, 3601, 151
https://tests.stockfishchess.org/tests/view/
60cefd1d457376eb8bcab7ab
LTC:
LLR: 2.94 (-2.94,2.94) <0.50,3.50>
Total: 10888 W: 386 L: 288 D: 10214
Ptnml(0-2): 3, 260, 4821, 356, 4
https://tests.stockfishchess.org/tests/view/
60cf709d2114332881e7352b
closes https://github.com/official-stockfish/Stockfish/pull/3571
Bench: 4965430
Fabian Fichter [Sun, 20 Jun 2021 09:15:25 +0000]
Replace CI badge
MichaelB7 [Sat, 19 Jun 2021 13:57:09 +0000]
Make net nn-
75980ca503c6.nnue the default.
trained with the Python command
c:\nnue>python train.py i:/bin/all.binpack i:/bin/all.binpack --gpus 1 --threads 4 --num-workers 30 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 3 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed %random%%random% --default_root_dir exp/run_10 --resume-from-model ./pt/nn-
3b20abec10c1.pt
`
all.binpack equaled 4 parts Wrong_NNUE_2.binpack https://drive.google.com/file/d/1seGNOqcVdvK_vPNq98j-zV3XPE5zWAeq/view?usp=sharing plus two parts of Training_Data.binpack https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing
Each set was concatenated together - making one large Wrong_NNUE 2 binpack and one large Training so the were approximately equal in size. They were then interleaved together. The idea was to give Wrong_NNUE.binpack closer to equal weighting with the Training_Data binpack .
Net nn-
3b20abec10c1.nnue was chosen as the --resume-from-model with the idea that through learning, the manually hex edited values will be learned and will not need to be manually adjusted going forward. They would also be fine tuned by the learning process.
passed STC:
https://tests.stockfishchess.org/tests/view/
60cdf91e457376eb8bcab66f
LLR: 2.95 (-2.94,2.94) <-0.50,2.50>
Total: 18256 W: 1639 L: 1479 D: 15138
Ptnml(0-2): 59, 1179, 6505, 1313, 72
passed LTC:
https://tests.stockfishchess.org/tests/view/
60ce2166457376eb8bcab6e1
LLR: 2.94 (-2.94,2.94) <0.50,3.50>
Total: 18792 W: 654 L: 542 D: 17596
Ptnml(0-2): 9, 490, 8291, 592, 14
closes https://github.com/official-stockfish/Stockfish/pull/3570
Bench: 5020972