This patch does similar thing to how it's done for
qsearch - in case of fail high adjust result to
lower value. Difference is that it is done only
for non-pv nodes and it's depth dependent - so
lower depth entries will have bigger adjustment
and higher depth entries will have smaller
adjustment.
Passed STC:
https://tests.stockfishchess.org/tests/view/65c3c0cbc865510db0283b21
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 112032 W: 29142 L: 28705 D: 54185
Ptnml(0-2): 479, 13152, 28326, 13571, 488
Passed LTC:
https://tests.stockfishchess.org/tests/view/65c52e62c865510db02855d5
LLR: 2.96 (-2.94,2.94) <0.50,2.50>
Total: 132480 W: 33457 L: 32936 D: 66087
Ptnml(0-2): 67, 14697, 36222, 15156, 98
closes https://github.com/official-stockfish/Stockfish/pull/5047
Bench: 1168241
In both search and qsearch, there are instances
where we do unadjustedStaticEval = ss->staticEval
= eval/bestValue = tte->eval(), but immediately
after re-assign ss-static and eval/bestValue to
some new value, which makes the initial assignment
redundant.
closes https://github.com/official-stockfish/Stockfish/pull/5045
No functional change
- Update codeql to v3
- Switch from dev-drprasad to native github cli
- Update softprops/action-gh-release to node 20 commit
`thollander/actions-comment-pull-request` needs to
be bumped to node20 too, but the author hasnt done
so atm
closes https://github.com/official-stockfish/Stockfish/pull/5044
No functional change
Move divisor from capture scoring to good capture
check and sligthly increase it.
This has several effects:
- its a speedup because for quience and probcut
search the division now never happens. For main
search its delayed and can be avoided if a good
capture triggers a cutoff
- through the higher resolution of scores we have
a more granular sorting
STC: https://tests.stockfishchess.org/tests/view/65bf2a93c865510db027dc27
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 470016 W: 122150 L: 121173 D: 226693
Ptnml(0-2): 2133, 55705, 118374, 56644, 2152
LTC: https://tests.stockfishchess.org/tests/view/65c1d16dc865510db0281339
LLR: 2.96 (-2.94,2.94) <0.50,2.50>
Total: 98988 W: 25121 L: 24667 D: 49200
Ptnml(0-2): 77, 10998, 26884, 11464, 71
closes https://github.com/official-stockfish/Stockfish/pull/5036
Bench: 1233867
This refactors the CI workflows to group some
logic and makes sure that all (pre)release
binaries are actually tested.
The screenshot below shows the execution logic of
the reworked ci,
https://github.com/Disservin/Stockfish/actions/runs/7773581379.
You can also hover over the cards to see the
execution flow.
The `matrix.json` and `arm_matrix.json` define the
binaries which will be uploaded to GitHub.
Afterwards a matrix is created and each job
compiles a profile guided build for that arch and
uploads that as an artifact to GitHub. The
Binaries/ARM_Binaries workflow's are called when
the previous step has been completed, and uploads
all artifacts to the (pre)release.
This also fixes some indentations and renames the
workflows, see
https://github.com/official-stockfish/Stockfish/actions,
where every workflow is called `Stockfish` vs
https://github.com/Disservin/Stockfish/actions. It
also increases the parallel compilation used for
make from `-j2 to -j4`.
It now also prevents the prerelease action from
running on forks.
A test release can be viewed here
https://github.com/Disservin/Stockfish/releases.
closes https://github.com/official-stockfish/Stockfish/pull/5035
No functional change
This replaces the PvNode condition and tte Pv call previously with using
the precomputed ttPv, and also removes the multiplier of 2. This new
depth condition occurs with approximately equal frequency (47%) to the
old depth condition (measured when the other conditions in the if are
true), so non-linear scaling behaviour isn't expected.
Passed Non-Reg STC:
https://tests.stockfishchess.org/tests/view/65b0e132c865510db026da27
LLR: 2.97 (-2.94,2.94) <-1.75,0.25>
Total: 243232 W: 62432 L: 62437 D: 118363
Ptnml(0-2): 910, 28937, 61900, 28986, 883
Passed Non-Reg LTC:
https://tests.stockfishchess.org/tests/view/65b2053bc865510db026eea1
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 190596 W: 47666 L: 47618 D: 95312
Ptnml(0-2): 115, 21710, 51596, 21766, 111
closes https://github.com/official-stockfish/Stockfish/pull/5015
Bench: 1492957
This splits the logic of search and perft. Before, threads were started,
which then constructed a search object, which then started perft and
returned immediately. All of this is unnecessary, instead uci should
start perft right away.
closes https://github.com/official-stockfish/Stockfish/pull/5008
No functional change
Update the internal WDL model. After the dual net merge, the internal
evaluations have drifted upwards a bit. With this PR
`NormalizeToPawnValue` changes from `328` to `345`.
The new model was fitted based on about 200M positions extracted from
3.4M fishtest LTC games from the last two weeks, involving SF versions
from 6deb88728f to current master.
Apart from the WDL model parameter update, this PR implements the
following changes:
WDL Model:
- an incorrect 8-move shift in master's WDL model has been fixed
- the polynomials `p_a` and `p_b` are fitted over the move range [8, 120]
- the coefficients for `p_a` and `p_b` are optimized by maximizing the
probability of predicting the observed outcome (credits to @vondele)
SF code:
- for wdl values, move will be clamped to `max(8, min(120, move))`
- no longer clamp the internal eval to [-4000,4000]
- compute `NormalizeToPawnValue` with `round`, not `trunc`
The PR only affects displayed `cp` and `wdl` values.
closes https://github.com/official-stockfish/Stockfish/pull/5002
No functional change
The idea of this is to unroll the futility_margin calculation to allow
for the improving flag to have a greater effect on the futility margin.
The current factor is 1.5 instead of the previous 1 resulting in a
deduction of an extra margin/2 from futilit_margin if improving. The
chosen value was not tuned, meaning that there is room for tweaking it.
This patch is partially inspired by @Vizvezdenec, who, although quite
different in execution, tested another idea where the futility_margin is
lowered further when improving [1].
[1]: (first take) https://tests.stockfishchess.org/tests/view/65a56b1879aa8af82b97164b
Passed STC:
https://tests.stockfishchess.org/tests/live_elo/65a8bfc179aa8af82b974e3c
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 161152 W: 41321 L: 40816 D: 79015
Ptnml(0-2): 559, 19030, 40921, 19479, 587
Passed rebased LTC:
https://tests.stockfishchess.org/tests/live_elo/65a8b9ef79aa8af82b974dc0
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 96024 W: 24172 L: 23728 D: 48124
Ptnml(0-2): 56, 10598, 26275, 11012, 71
closes https://github.com/official-stockfish/Stockfish/pull/5000
Bench: 1281703
This addresses the issue where Stockfish may output non-proven checkmate
scores if the search is prematurely halted, either due to a time control
or node limit, before it explores other possibilities where the
checkmate score could have been delayed or refuted.
The fix also replaces staving off from proven mated scores in a
multithread environment making use of the threads instead of a negative
effect with multithreads (1t was better in proving mated in scores than
more threads).
Issue reported on mate tracker repo by and this PR is co-authored with
@robertnurnberg Special thanks to @AndyGrant for outlining that a fix is
eventually possible.
Passed Adj off SMP STC:
https://tests.stockfishchess.org/tests/view/65a125d779aa8af82b96c3eb
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 303256 W: 75823 L: 75892 D: 151541
Ptnml(0-2): 406, 35269, 80395, 35104, 454
Passed Adj off SMP LTC:
https://tests.stockfishchess.org/tests/view/65a37add79aa8af82b96f0f7
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 56056 W: 13951 L: 13770 D: 28335
Ptnml(0-2): 11, 5910, 16002, 6097, 8
Passed all tests in matetrack without any better mate for opponent found in 1t and multithreads.
Fixed bugs in https://github.com/official-stockfish/Stockfish/pull/4976
closes https://github.com/official-stockfish/Stockfish/pull/4990
Bench: 1308279
Co-Authored-By: Robert Nürnberg <28635489+robertnurnberg@users.noreply.github.com>
Also remove dead code, `rootSimpleEval` is no longer used since the introduction of dual net.
`iterBestValue` is also no longer used in evaluate and can be reduced to a local variable.
closes https://github.com/official-stockfish/Stockfish/pull/4979
No functional change
This aims to remove some of the annoying global structure which Stockfish has.
Overall there is no major elo regression to be expected.
Non regression SMP STC (paused, early version):
https://tests.stockfishchess.org/tests/view/65983d7979aa8af82b9608f1
LLR: 0.23 (-2.94,2.94) <-1.75,0.25>
Total: 76232 W: 19035 L: 19096 D: 38101
Ptnml(0-2): 92, 8735, 20515, 8690, 84
Non regression STC (early version):
https://tests.stockfishchess.org/tests/view/6595b3a479aa8af82b95da7f
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 185344 W: 47027 L: 46972 D: 91345
Ptnml(0-2): 571, 21285, 48943, 21264, 609
Non regression SMP STC:
https://tests.stockfishchess.org/tests/view/65a0715c79aa8af82b96b7e4
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 142936 W: 35761 L: 35662 D: 71513
Ptnml(0-2): 209, 16400, 38135, 16531, 193
These global structures/variables add hidden dependencies and allow data
to be mutable from where it shouldn't it be (i.e. options). They also
prevent Stockfish from internal selfplay, which would be a nice thing to
be able to do, i.e. instantiate two Stockfish instances and let them
play against each other. It will also allow us to make Stockfish a
library, which can be easier used on other platforms.
For consistency with the old search code, `thisThread` has been kept,
even though it is not strictly necessary anymore. This the first major
refactor of this kind (in recent time), and future changes are required,
to achieve the previously described goals. This includes cleaning up the
dependencies, transforming the network to be self contained and coming
up with a plan to deal with proper tablebase memory management (see
comments for more information on this).
The removal of these global structures has been discussed in parts with
Vondele and Sopel.
closes https://github.com/official-stockfish/Stockfish/pull/4968
No functional change
Created by retraining the previous main net nn-b1e55edbea57.nnue with:
- some of the same options as before: ranger21 optimizer, more WDL
skipping
- adding T80 aug filter-v6, sep, and oct 2023 data to the previous best
dataset
- increasing training loss for positions where predicted win rates were
higher than estimated match results from training data position scores
```yaml
experiment-name: 2560--S8-r21-more-wdl-skip-10p-more-loss-high-q-sk28
training-dataset:
# https://github.com/official-stockfish/Stockfish/pull/4782
- /data/S6-1ee1aba5ed.binpack
- /data/test80-aug2023-2tb7p.v6.min.binpack
- /data/test80-sep2023-2tb7p.binpack
- /data/test80-oct2023-2tb7p.binpack
early-fen-skipping: 28
start-from-engine-test-net: True
nnue-pytorch-branch: linrock/nnue-pytorch/r21-more-wdl-skip-10p-more-loss-high-q
num-epochs: 1000
lr: 4.375e-4
gamma: 0.995
start-lambda: 1.0
end-lambda: 0.7
```
Training data can be found at:
https://robotmoon.com/nnue-training-data/
Training loss was increased by 10% for positions where predicted win
rates were higher than suggested by the win rate model based on the
training data, by multiplying with: ((qf > pt) * 0.1 + 1). This was a
variant of experiments from Sopel's NNUE training & experimentation log:
https://docs.google.com/document/d/1gTlrr02qSNKiXNZ_SuO4-RjK4MXBiFlLE6jvNqqMkAY
Experiment 302 - increase loss when prediction too high, vondele’s idea
Experiment 309 - increase loss when prediction too high, normalize in a
batch
Passed STC:
https://tests.stockfishchess.org/tests/view/6597a21c79aa8af82b95fd5c
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 148320 W: 37960 L: 37475 D: 72885
Ptnml(0-2): 542, 17565, 37383, 18206, 464
Passed LTC:
https://tests.stockfishchess.org/tests/view/659834a679aa8af82b960845
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 55188 W: 13955 L: 13592 D: 27641
Ptnml(0-2): 34, 6162, 14834, 6535, 29
closes https://github.com/official-stockfish/Stockfish/pull/4972
Bench: 1219824
Created by training an L1-128 net from scratch with a wider range of
evals in the training data and wld-fen-skipping disabled during
training. The differences in this training data compared to the first
dual nnue PR are:
- removal of all positions with 3 pieces
- when piece count >= 16, keep positions with simple eval above 750
- when piece count < 16, remove positions with simple eval above 3000
The asymmetric data filtering was meant to flatten the training data
piece count distribution, which was previously heavily skewed towards
positions with low piece counts.
Additionally, the simple eval range where the smallnet is used was
widened to cover more positions previously evaluated by the big net and
simple eval.
```yaml
experiment-name: 128--S1-hse-S7-v4-S3-v1-no-wld-skip
training-dataset:
- /data/hse/S3/leela96-filt-v2.min.high-simple-eval-1k.binpack
- /data/hse/S3/dfrc99-16tb7p-eval-filt-v2.min.high-simple-eval-1k.binpack
- /data/hse/S3/test80-apr2022-16tb7p.min.high-simple-eval-1k.binpack
- /data/hse/S7/test60-2020-2tb7p.v6-3072.high-simple-eval-v4.binpack
- /data/hse/S7/test60-novdec2021-12tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test77-nov2021-2tb7p.v6-3072.min.high-simple-eval-v4.binpack
- /data/hse/S7/test77-dec2021-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test77-jan2022-2tb7p.high-simple-eval-v4.binpack
- /data/hse/S7/test78-jantomay2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test78-juntosep2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test79-apr2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test79-may2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-may2022-16tb7p.high-simple-eval-v4.binpack
- /data/hse/S7/test80-jun2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-jul2022-16tb7p.v6-dd.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-aug2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-sep2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-oct2022-16tb7p.v6-dd.high-simple-eval-v4.binpack
- /data/hse/S7/test80-nov2022-16tb7p-v6-dd.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-jan2023-3of3-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-feb2023-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-v4.binpack
- /data/hse/S7/test80-mar2023-2tb7p.v6-sk16.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-apr2023-2tb7p-filter-v6-sk16.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-may2023-2tb7p.v6.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-jun2023-2tb7p.v6-3072.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-jul2023-2tb7p.v6-3072.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-aug2023-2tb7p.v6.min.high-simple-eval-v4.binpack
- /data/hse/S7/test80-sep2023-2tb7p.high-simple-eval-v4.binpack
- /data/hse/S7/test80-oct2023-2tb7p.high-simple-eval-v4.binpack
wld-fen-skipping: False
start-from-engine-test-net: False
nnue-pytorch-branch: linrock/nnue-pytorch/L1-128
engine-test-branch: linrock/Stockfish/L1-128-nolazy
engine-base-branch: linrock/Stockfish/L1-128
num-epochs: 500
start-lambda: 1.0
end-lambda: 1.0
```
Experiment yaml configs converted to easy_train.sh commands with:
https://github.com/linrock/nnue-tools/blob/4339954/yaml_easy_train.py
Binpacks interleaved at training time with:
https://github.com/official-stockfish/nnue-pytorch/pull/259
FT weights permuted with 10k positions from fishpack32.binpack with:
https://github.com/official-stockfish/nnue-pytorch/pull/254
Data filtered for high simple eval positions (v4) with:
https://github.com/linrock/Stockfish/blob/b9c8440/src/tools/transform.cpp#L640-L675
Training data can be found at:
https://robotmoon.com/nnue-training-data/
Local elo at 25k nodes per move of
L1-128 smallnet (nnue-only eval) vs. L1-128 trained on standard S1 data:
nn-epoch319.nnue : -241.7 +/- 3.2
Passed STC vs. 36db936:
https://tests.stockfishchess.org/tests/view/6576b3484d789acf40aabbfe
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 21920 W: 5680 L: 5381 D: 10859
Ptnml(0-2): 82, 2488, 5520, 2789, 81
Passed LTC vs. DualNNUE #4915:
https://tests.stockfishchess.org/tests/view/65775c034d789acf40aac7e3
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 147606 W: 36619 L: 36063 D: 74924
Ptnml(0-2): 98, 16591, 39891, 17103, 120
closes https://github.com/official-stockfish/Stockfish/pull/4919
Bench: 1438336
