The idea of this patch can be described as following: we update static
history stats based on comparison of the static evaluations of the
position before and after the move. If the move increases static evaluation
it's assigned positive bonus, if it decreases static evaluation
it's assigned negative bonus. These stats are used in movepicker
to sort quiet moves.
passed STC
https://tests.stockfishchess.org/tests/view/5fca4c0842a050a89f02cd66
LLR: 3.00 (-2.94,2.94) {-0.25,1.25}
Total: 78152 W: 7409 L: 7171 D: 63572
Ptnml(0-2): 303, 5695, 26873, 5871, 334
passed LTC
https://tests.stockfishchess.org/tests/view/5fca6be442a050a89f02cd75
LLR: 2.94 (-2.94,2.94) {0.25,1.25}
Total: 40240 W: 1602 L: 1441 D: 37197
Ptnml(0-2): 19, 1306, 17305, 1475, 15
closes https://github.com/official-stockfish/Stockfish/pull/3253
bench 3845156
Idea of this patch can be described as following - in case we have consecutive fail highs and we reach late enough moves at root node probability of remaining quiet moves being able to produce even bigger value than moves that produced previous cutoff (so ones that should be high in move ordering but now they fail to produce beta cutoff because we actually reached high move count) should be quiet small so we can reduce them more.
passed STC
LLR: 2.94 (-2.94,2.94) {-0.25,1.25}
Total: 53392 W: 5681 L: 5474 D: 42237
Ptnml(0-2): 214, 4104, 17894, 4229, 255
https://tests.stockfishchess.org/tests/view/5f88501adcdad978fe8c527e
passed LTC
LLR: 2.94 (-2.94,2.94) {0.25,1.25}
Total: 59136 W: 2773 L: 2564 D: 53799
Ptnml(0-2): 30, 2117, 25078, 2300, 43
https://tests.stockfishchess.org/tests/view/5f884dbfdcdad978fe8c527a
closes https://github.com/official-stockfish/Stockfish/pull/3184
Bench: 4066972
RootInTB was an incorrectly shared global, probably leading to wrong scoreing
Minor:
setting TB global state from input by all threads (all threads write same values)
setting Limits global state by all threads (idem)
thread counting for finalization
CI can be enabled once races are fixed in the learner, manually goes like:
```
make clean && make -j2 ARCH=x86-64-modern sanitize=thread optimize=no debug=yes build
../tests/instrumented_learn.sh --sanitizer-thread
```
Needs some review.
This patch ports the efficiently updatable neural network (NNUE) evaluation to Stockfish.
Both the NNUE and the classical evaluations are available, and can be used to
assign a value to a position that is later used in alpha-beta (PVS) search to find the
best move. The classical evaluation computes this value as a function of various chess
concepts, handcrafted by experts, tested and tuned using fishtest. The NNUE evaluation
computes this value with a neural network based on basic inputs. The network is optimized
and trained on the evalutions of millions of positions at moderate search depth.
The NNUE evaluation was first introduced in shogi, and ported to Stockfish afterward.
It can be evaluated efficiently on CPUs, and exploits the fact that only parts
of the neural network need to be updated after a typical chess move.
[The nodchip repository](https://github.com/nodchip/Stockfish) provides additional
tools to train and develop the NNUE networks.
This patch is the result of contributions of various authors, from various communities,
including: nodchip, ynasu87, yaneurao (initial port and NNUE authors), domschl, FireFather,
rqs, xXH4CKST3RXx, tttak, zz4032, joergoster, mstembera, nguyenpham, erbsenzaehler,
dorzechowski, and vondele.
This new evaluation needed various changes to fishtest and the corresponding infrastructure,
for which tomtor, ppigazzini, noobpwnftw, daylen, and vondele are gratefully acknowledged.
The first networks have been provided by gekkehenker and sergiovieri, with the latter
net (nn-97f742aaefcd.nnue) being the current default.
The evaluation function can be selected at run time with the `Use NNUE` (true/false) UCI option,
provided the `EvalFile` option points the the network file (depending on the GUI, with full path).
The performance of the NNUE evaluation relative to the classical evaluation depends somewhat on
the hardware, and is expected to improve quickly, but is currently on > 80 Elo on fishtest:
60000 @ 10+0.1 th 1
https://tests.stockfishchess.org/tests/view/5f28fe6ea5abc164f05e4c4c
ELO: 92.77 +-2.1 (95%) LOS: 100.0%
Total: 60000 W: 24193 L: 8543 D: 27264
Ptnml(0-2): 609, 3850, 9708, 10948, 4885
40000 @ 20+0.2 th 8
https://tests.stockfishchess.org/tests/view/5f290229a5abc164f05e4c58
ELO: 89.47 +-2.0 (95%) LOS: 100.0%
Total: 40000 W: 12756 L: 2677 D: 24567
Ptnml(0-2): 74, 1583, 8550, 7776, 2017
At the same time, the impact on the classical evaluation remains minimal, causing no significant
regression:
sprt @ 10+0.1 th 1
https://tests.stockfishchess.org/tests/view/5f2906a2a5abc164f05e4c5b
LLR: 2.94 (-2.94,2.94) {-6.00,-4.00}
Total: 34936 W: 6502 L: 6825 D: 21609
Ptnml(0-2): 571, 4082, 8434, 3861, 520
sprt @ 60+0.6 th 1
https://tests.stockfishchess.org/tests/view/5f2906cfa5abc164f05e4c5d
LLR: 2.93 (-2.94,2.94) {-6.00,-4.00}
Total: 10088 W: 1232 L: 1265 D: 7591
Ptnml(0-2): 49, 914, 3170, 843, 68
The needed networks can be found at https://tests.stockfishchess.org/nns
It is recommended to use the default one as indicated by the `EvalFile` UCI option.
Guidelines for testing new nets can be found at
https://github.com/glinscott/fishtest/wiki/Creating-my-first-test#nnue-net-tests
Integration has been discussed in various issues:
https://github.com/official-stockfish/Stockfish/issues/2823https://github.com/official-stockfish/Stockfish/issues/2728
The integration branch will be closed after the merge:
https://github.com/official-stockfish/Stockfish/pull/2825https://github.com/official-stockfish/Stockfish/tree/nnue-player-wip
closes https://github.com/official-stockfish/Stockfish/pull/2912
This will be an exciting time for computer chess, looking forward to seeing the evolution of
this approach.
Bench: 4746616
Current move histories are known to work well near the leaves, whilst at
higher depths they aren't very helpful. To address this problem this
patch introduces a table dedicated for what's happening at plies 0-3.
It's structured like mainHistory with ply index instead of color.
It get cleared with each new search and is filled during iterative
deepening at higher depths when recording successful quiet moves near
the root or traversing nodes which were in the principal variation
(ttPv).
Medium TC (20+0.2):
https://tests.stockfishchess.org/tests/view/5e4d358790a0a02810d096dc
LLR: 2.94 (-2.94,2.94) {-0.50,1.50}
Total: 100910 W: 16682 L: 16376 D: 67852
Ptnml(0-2): 1177, 10983, 25883, 11181, 1231
LTC:
https://tests.stockfishchess.org/tests/view/5e4e2cb790a0a02810d09714
LLR: 2.95 (-2.94,2.94) {0.25,1.75}
Total: 80444 W: 10495 L: 10095 D: 59854
Ptnml(0-2): 551, 7479, 23803, 7797, 592
closes https://github.com/official-stockfish/Stockfish/pull/2557
Bench: 4705960
This patch makes Stockfish search same depth again if > 60% of optimum time is
already used, instead of trying the next iteration. The idea is that the next
iteration will generally take about the same amount of time as has already been
used in total. When we are likely to begin the last iteration, as judged by total
time taken so far > 0.6 * optimum time, searching the last depth again instead of
increasing the depth still helps the other threads in lazy SMP and prepares better
move ordering for the next moves.
STC :
LLR: 2.95 (-2.94,2.94) {-1.00,3.00}
Total: 13436 W: 2695 L: 2558 D: 8183
Ptnml(0-2): 222, 1538, 3087, 1611, 253
https://tests.stockfishchess.org/tests/view/5e1618a761fe5f83a67dd964
LTC :
LLR: 2.94 (-2.94,2.94) {0.00,2.00}
Total: 32160 W: 4261 L: 4047 D: 23852
Ptnml(0-2): 211, 2988, 9448, 3135, 247
https://tests.stockfishchess.org/tests/view/5e162ca061fe5f83a67dd96d
The code was revised as suggested by @vondele for multithreading:
STC (8 threads):
LLR: 2.95 (-2.94,2.94) {0.00,2.00}
Total: 16640 W: 2049 L: 1885 D: 12706
Ptnml(0-2): 119, 1369, 5158, 1557, 108
https://tests.stockfishchess.org/tests/view/5e19826a2cc590e03c3c2f52
LTC (8 threads):
LLR: 2.95 (-2.94,2.94) {-1.00,3.00}
Total: 16536 W: 2758 L: 2629 D: 11149
Ptnml(0-2): 182, 1758, 4296, 1802, 224
https://tests.stockfishchess.org/tests/view/5e18b91a27dab692fcf9a140
Thanks to those discussing Stockfish lazy SMP on fishcooking which made me
try this, and to @vondele for suggestions and doing related tests.
See full discussion in the pull request thread:
https://github.com/official-stockfish/Stockfish/pull/2482
Bench: 4586187
This patch measures how frequently search is exploring new configurations.
This is done be computing a running average of ttHit. The ttHitAverage rate
is somewhat low (e.g. 30% for startpos) in the normal case, while it can be
very high if no progress is made (e.g. 90% for the fortress I used for testing).
This information can be used to influence search. In this patch, by adjusting
reductions if the rate > 50%. A first version (using a low ttHitAverageResolution
and this 50% threshold) passed testing:
STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 26425 W: 5837 L: 5650 D: 14938
http://tests.stockfishchess.org/tests/view/5dcede8b0ebc5902563258fa
LTC
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 32313 W: 5392 L: 5128 D: 21793
http://tests.stockfishchess.org/tests/view/5dcefb1f0ebc590256325c0e
However, as discussed in pull request 2414, using a larger ttHitAverageResolution
gives a better approximation of the underlying distributions. This needs a slight
adjustment for the threshold as the new distributions are shifted a bit compared
to the older ones, and this threshold seemingly is sensitive (we used 0.53125 here).
https://github.com/official-stockfish/Stockfish/pull/2414
This final version also passed testing, and is used for the patch:
STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 16025 W: 3555 L: 3399 D: 9071
http://tests.stockfishchess.org/tests/view/5dd070b90ebc5902579e20c2
LTC
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 37576 W: 6277 L: 5998 D: 25301
http://tests.stockfishchess.org/tests/view/5dd0f58e6f544e798086f224
Closes https://github.com/official-stockfish/Stockfish/pull/2414
Bench: 4989584
Maintain best move counter at the root and allow there only moves which has a counter
of zero for Late Move Reduction. For compensation only the first three moves are excluded
from Late Move Reduction per default instead the first four moves.
What we can further do:
- here we use a simple counting scheme but perhaps some aging to fade out early iterations
could be helpful
- use the best move counter also at inner nodes for LMR and/or pruning
STC:
LLR: 2.95 (-2.94,2.94) [0.50,4.50]
Total: 17414 W: 3984 L: 3733 D: 9697
http://tests.stockfishchess.org/tests/view/5d6234bb0ebc5939d09f2aa2
LTC:
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 38058 W: 6448 L: 6166 D: 25444
http://tests.stockfishchess.org/tests/view/5d62681a0ebc5939d09f2f27
Closes https://github.com/official-stockfish/Stockfish/pull/2282
Bench: 3568210
Fixes issues #2126 and #2189 where no progress in rootDepth is made for particular fens:
8/8/3P3k/8/1p6/8/1P6/1K3n2 b - - 0 1
8/1r1rp1k1/1b1pPp2/2pP1Pp1/1pP3Pp/pP5P/P5K1/8 w - - 79 46
the cause are the shuffle extensions. Upon closer analysis, it appears that in these cases a shuffle extension is made for every node searched, and progess can not be made. This patch implements a fix, namely to limit the number of extensions relative to the number of nodes searched. The ratio employed is 1/4, which fixes the issues seen so far, but it is a heuristic, and I expect that certain positions might require an even smaller fraction.
The patch was tested as a bug fix and passed:
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 56601 W: 12633 L: 12581 D: 31387
http://tests.stockfishchess.org/tests/view/5d02b37a0ebc5925cf09f6da
LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 52042 W: 8907 L: 8837 D: 34298
http://tests.stockfishchess.org/tests/view/5d0319420ebc5925cf09fe57
Furthermore, to confirm that the shuffle extension in this form indeed still brings Elo, one more test at VLTC was performed:
VLTC:
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 142022 W: 20963 L: 20435 D: 100624
http://tests.stockfishchess.org/tests/view/5d03630d0ebc5925cf0a011a
Bench: 3961247
Similar to PSQT we only need one instance of the Endgames resource. The current per thread copies are identical and read only(after initialization) so from a design point of view it doesn't make sense to have them.
Tested for no slowdown.
http://tests.stockfishchess.org/tests/view/5c94377a0ebc5925cfff43ca
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17320 W: 3487 L: 3359 D: 10474
No functional change.
The current update only by main thread depends on the luck of
whether main thread sees any/many changes to the best move or not.
It then makes large, lumpy changes to the time to be
used (1x, 2x, 3x, etc) depending on that sample of 1.
Use the average across all threads to get a more reliable
number with a smoother distribution.
STC @ 5+0.05 th 4 :
LLR: 2.95 (-2.94,2.94) [0.50,4.50]
Total: 51899 W: 11446 L: 11029 D: 29424
http://tests.stockfishchess.org/tests/view/5ca32ff20ebc5925cf0016fb
STC @ 5+0.05 th 8 :
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 13851 W: 2843 L: 2620 D: 8388
http://tests.stockfishchess.org/tests/view/5ca35ae00ebc5925cf001adb
LTC @ 20+0.2 th 8 :
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 48527 W: 7941 L: 7635 D: 32951
http://tests.stockfishchess.org/tests/view/5ca37cb70ebc5925cf001cec
Further work:
Similar changes might be possible for the fallingEval and timeReduction calculations (and elsewhere?), using either the min, average or max values across all threads.
Bench 3506898
On OS X threads other than the main thread are created with a reduced stack
size of 512KB by default, this is dangerously low for deep searches, so
adjust it to TH_STACK_SIZE. The implementation calls pthread_create() with
proper stack size parameter.
Verified for no regression at STC enabling the patch on all platforms where
pthread is supported.
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 50873 W: 9768 L: 9700 D: 31405
No functional change.
stopOnPonderhit is used to stop search quickly on a ponderhit. It is set by mainThread as part of its time management. However, master employs it as a signal between mainThread and the UCI thread. This is not necessary, it is sufficient for the UCI thread to signal that pondering finished, and mainThread should do its usual time-keeping job, and in this case stop immediately.
This patch implements this, removing stopOnPonderHit as an atomic variable from the ThreadPool,
and moving it as a normal variable to mainThread, reducing its scope. In MainThread::check_time() the search is stopped immediately if ponder switches to false, and the variable stopOnPonderHit is set.
Furthermore, ponder has been moved to mainThread, as the variable is only used to exchange signals between the UCI thread and mainThread.
The version has been tested locally (as fishtest doesn't support ponder):
Score of ponderSimp vs master: 2616 - 2528 - 8630 [0.503] 13774
Elo difference: 2.22 +/- 3.54
which indicates no regression.
No functional change.
Preparation commit for the upcoming Stockfish 10 version, giving a chance to catch last minute feature bugs and evaluation regression during the one-week code freeze period. Also changing the copyright dates to include 2019.
No functional change
define Color us and use this instead of pos.side_to_move() and nmp_odd. The latter allows to clarify the nmp verification criterion.
Tested for no regression:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 76713 W: 15303 L: 15284 D: 46126
http://tests.stockfishchess.org/tests/view/5b046a0d0ebc5914abc12971
No functional change.
This patch corrects both MultiPV behaviour and "go searchmoves" behaviour
for tablebases.
We change the logic of table base probing at root positions from filtering
to ranking. The ranking code is much more straightforward than the current
filtering code (this is a simplification), and also more versatile.
If the root is a TB position, each root move is probed and assigned a TB score
and a TB rank. The TB score is the Value to be displayed to the user for that
move (unless the search finds a mate score), while the TB rank determines which
moves should appear higher in a multi-pv search. In game play, the engine will
always pick a move with the highest rank.
Ranks run from -1000 to +1000:
901 to 1000 : TB win
900 : normally a TB win, in rare cases this could be a draw
1 to 899 : cursed TB wins
0 : draw
-1 to -899 : blessed TB losses
-900 : normally a TB loss, in rare cases this could be a draw
-901 to -1000 : TB loss
Normally all winning moves get rank 1000 (to let the search pick the best
among them). The exception is if there has been a first repetition. In that
case, moves are ranked strictly by DTZ so that the engine will play a move
that lowers DTZ (and therefore cannot repeat the position a second time).
Losing moves get rank -1000 unless they have relatively high DTZ, meaning
they have some drawing chances. Those get ranks towards -901 (when they
cross -900 the draw is certain).
Closes https://github.com/official-stockfish/Stockfish/pull/1467
No functional change (without tablebases).
