TOSHIBA LTC CHESS ENGINE TOURNAMENT

[Engine]

UPCOMING TOURNAMENTE ENGINE

ENGINE INFO
SFT 230325-0d8c4-587bc 587bc647-1, SEE and promotions, Bench: 5420267, MinetaS committed Mar 25, 2023, 1 parent 587bc64 commit 0d8c43f, Elo gain: 2.21 [0.62,3.79] (95%). By the Stockfish developers (see AUTHORS file)


SPEED TEST (BENCH COMAND, INTEL I3 2.00 GHZ, BMI2-PGO)
Total time (ms) : 21857
Nodes searched : 5420267
Nodes/second : 247987


DOWNLOAD LINK:
https://www.file-upload.com/zswkgviw5tj2

[Engine]

Author: Linmiao Xu
Date: Wed Mar 29 21:37:52 2023 +0200
Timestamp: 1680118672

Update default net to nn-dabb1ed23026.nnue

Created by retraining the master net with these modifications:

* New filtering methods for existing data from T80 sep+oct2022, T79 apr2022, T78 jun+jul+aug+sep2022, T77 dec2021
* Adding new filtered data from T80 aug2022 and T78 apr+may2022
* Increasing early-fen-skipping from 28 to 30

```
python3 easy_train.py \
--experiment-name leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3-sk30 \
--training-dataset /data/leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3.binpack \
--nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes \
--start-from-engine-test-net True \
--early-fen-skipping 30 \
--max_epoch 900 \
--start-lambda 1.0 \
--end-lambda 0.7 \
--lr 4.375e-4 \
--gamma 0.995 \
--tui False \
--gpus "0," \
--seed $RANDOM
```

The v3 filtering used for data from T77dec 2021 differs from v2 filtering in that:

* To improve binpack compression, positions after ply 28 were skipped during training by setting position scores to VALUE_NONE (32002) instead of removing them entirely
* All early-game positions with ply <= 28 were removed to maximize binpack compression
* Only bestmove captures at d6pv2 search were skipped, not 2nd bestmove captures
* Binpack compression was repaired for the remaining positions by effectively replacing bestmoves with "played moves" to maintain contiguous sequences of positions in the training game data

After improving binpack compression, The T77 dec2021 data size was reduced from 95G to 19G.

The v6 filtering used for data from T80augsepoctT79aprT78aprtosep 2022 differs from v2 in that:

* All positions with only one legal move were removed
* Tighter score differences at d6pv2 search were used to remove more positions with only one good move than before
* d6pv2 search was not used to remove positions where the best 2 moves were captures

```
python3 interleave_binpacks.py \
nn-547-dataset/leela96-eval-filt-v2.binpack \
nn-547-dataset/dfrc99-eval-filt-v2.binpack \
nn-547-dataset/test80-nov2022-12tb7p-eval-filt-v2-d6.binpack \
nn-547-dataset/T79-may2022-12tb7p-eval-filt-v2.binpack \
nn-547-dataset/T60-nov2021-12tb7p-eval-filt-v2.binpack \
nn-547-dataset/T60-dec2021-12tb7p-eval-filt-v2.binpack \
filt-v6/test80-aug2022-16tb7p-filter-v6.binpack \
filt-v6/test80-sep2022-16tb7p-filter-v6.binpack \
filt-v6/test80-oct2022-16tb7p-filter-v6.binpack \
filt-v6/test79-apr2022-16tb7p-filter-v6.binpack \
filt-v6/test78-aprmay2022-16tb7p-filter-v6.binpack \
filt-v6/test78-junjulaug2022-16tb7p-filter-v6.binpack \
filt-v6/test78-sep2022-16tb7p-filter-v6.binpack \
filt-v3/test77-dec2021-16tb7p-filt-v3.binpack \
/data/leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3.binpack
```

The code for the new data filtering methods is available at:
https://github.com/linrock/Stockfish/tr ... /nnue-data

The code for giving hexword names to .nnue files is at:
https://github.com/linrock/nnue-namer

Links for downloading the training data components can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch779.nnue : 0.6 +/- 3.1

Passed STC:
https://tests.stockfishchess.org/tests/ ... 2ba6f8efb0
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 82256 W: 22185 L: 21809 D: 38262 Elo +1.59
Ptnml(0-2): 286, 9065, 22067, 9407, 303

Passed LTC:
https://tests.stockfishchess.org/tests/ ... 2ba6f91d6c
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 30840 W: 8437 L: 8149 D: 14254 Elo +3.24
Ptnml(0-2): 14, 2891, 9323, 3177, 15

closes https://github.com/official-stockfish/S ... /pull/4465

bench 5101970

https://abrok.eu/stockfish/

[mig2004]

UPCOMING ENGINE

ENGINE NAME
SFT 230328-f7e60-587bc- master (Remove non_pawn_material in NNUEevaluate). This maybe the same as the previous post from abrok.eu At any rate, i believe the mingw compiler, the one i use, produces a faster executable. I haven´t retested them for 5 years now, it may have changed somewhat).

ENGINE INFO
SFT 230328-f7e60-587bc 587bc647-4, Attempt 3: Limit complexity update, second variant, Bench: 5020718, 587bc647-4 , MinetaS committed Mar 28, 2023, 1 parent 8bbfc1c commit f7e6001, Elo gain: 3.04 [1.15,4.91] (95%) over master 587bc- (Remove non_pawn_material in NNUE evaluate), by the Stockfish developers (see AUTHORS file)
ENGINE SPEED TEST (INTEL I3, 2 GHZ, BMI2-PGO,)
Total time (ms) : 20700
Nodes searched : 5020718
Nodes/second : 242546

DOWNLOAD LINK: https://www.file-upload.com/i737zf2l75g4

[mig2004]

NEW UPCOMING ENGINE

ENGINE NAME
SFT 230330-ffc66-3f01e - master (Allow PvNode in futility pruning for captures.)

ENGINE INFO

SFT 230330-ffc66-3f01e moreIIRa2b, Update search.cpp , moreIIRa2b , FauziAkram committed Mar 30, 2023 , 1 parent 3f01e3f commit ffc66db . Elo gain: 2.55 [0.82,4.26] (95%) over master 3f01e - (Allow PvNode in futility pruning for captures.). By the Stockfish developers (see AUTHORS file)


DOWNLOAD LINK: https://www.file-upload.com/ks6rvu013jxf

[mig2004]

UPCOMING TURNAMENT ENGINE

LAST OF THE SERIES 587bc- master (Remove non_pawn_material in NNUE::evaluate)
Lmr capture condition tuning: only parameter A[7][2] with rule random…compensation p=7/8. Bench: 4745336, tuned_lmr_cap2 , Stefan Geschwentner committed Mar 28, 2023, 1 parent c2562a0 commit 4b49bae.
ORIGINAL TEST INFO
23-03-28 sg tuned_lmr_cap2 diff LLR: 2.94 (-2.94,2.94) <0.00,2.00>Total: 34592 W: 9316 L: 9011 D: 16265Ptnml(0-2): 101, 3739, 9325, 4016, 115 sprt @ 10+0.1 th 1 Lmr capture condition tuning: only parameter A[7][2] with rule random compensation p=7/8. 587bc- master (Remove non_pawn_material in NNUE::evaluate) 2.90 [1.03,4.75] (95%)


ENGINE NAME
SFT 230328-4b49b-587bc- master (Remove non_pawn_material in NNUe evaluate)

ENGINE INFO
SFT 230328-4b49b-587bc ORIGIN NAME:tuned_lmr_cap2, COMMIT COMMENTS: Lmr capture condition tuning: only parameter A[7][2] with rule randomcompensation p=7/8. Bench: 4745336, tuned_lmr_cap2 , Stefan Geschwentner committed Mar 28, 2023, 1 parent c2562a0 commit 4b49bae. Elo gain: 2.90 [1.03,4.75] (95%) over maser. By the Stockfish developers (see AUTHORS file)

SPEED TEST (intel i3, 2.00 GHZ)
Total time (ms) : 19641
Nodes searched : 4745336
Nodes/second : 241603
DOWNLOAD LINIK: https://www.file-upload.com/d03av659k1q5

[mig2004]

***NEW ENGINE TO QUEUE
***TOURNAMENT RANKING AS OF APRIL 1, 2023
(copy and paste into your favorite word procesor-spreadsheet for better view)

TOP 50 PER ELO, MINIMUM 6 GAMES PLAYED.

Rank Name Rating Δ + - # Σ Σ% W L D W% =% OppR
---------------------------------------------------------------------------------------------------------
1 Stockfish_230308_T169 nn5247 3442 0.0 39 64 6 4.0 66.7 3 1 2 50.0 33.3 3378
2 SFT 230303-02931-87690_nn5247 3429 12.5 48 67 6 3.5 58.3 1 0 5 16.7 83.3 3394
3 SF Official 221016-9be2_nnad9 3429 0.3 41 50 10 6.5 65.0 3 0 7 30.0 70.0 3369
4 Stockfish_230305_T247 nn5247 3428 1.0 47 64 6 4.0 66.7 2 0 4 33.3 66.7 3363
5 SFT 230306-1ac75-cdec7_nn5247 3425 2.5 51 68 6 3.5 58.3 3 2 1 50.0 16.7 3391
6 Stockfish_230308_T047_nn5247 3425 0.6 51 67 6 3.5 58.3 2 1 3 33.3 50.0 3391
7 Stockfish_221202_T829 nnad9b 3419 5.6 47 51 10 6.0 60.0 4 2 4 40.0 40.0 3376
8 SFT 230308-5ccc8-39da5_nn5247 3419 0.5 53 65 6 3.5 58.3 2 1 3 33.3 50.0 3386
9 Stockfish_221105_T153 nnad9b 3414 4.9 44 49 12 7.0 58.3 3 1 8 25.0 66.7 3377
10 SF Official 221008-e903_nnad9 3412 1.2 35 40 18 10.5 58.3 4 1 13 22.2 72.2 3375
11 Stockfish_230308_T089 nn5247 3408 4.7 56 71 6 3.5 58.3 2 1 3 33.3 50.0 3370
12 Stockfish_230305_T312 nn5247 3403 4.9 62 64 6 3.0 50.0 1 1 4 16.7 66.7 3403
13 SF Official 221016-79c5_nnad9 3402 0.7 59 72 6 3.5 58.3 3 2 1 50.0 16.7 3362
14 SFT 221104-99f59-d0965_nnad9b 3400 2.1 45 47 12 6.5 54.2 3 2 7 25.0 58.3 3383
15 Stockfish_22100816_T250nnad9b 3400 0.3 60 72 6 3.5 58.3 3 2 1 50.0 16.7 3360
16 Stockfish_230305_763 nn5247 3398 2.3 62 63 6 3.0 50.0 0 0 6 0.0 100.0 3399
17 SFT 220914-C82B0 nnad9b 3397 0.4 38 42 16 9.0 56.2 2 0 14 12.5 87.5 3369
18 Stockfish_221105_T114 nnad9b 3388 9.2 67 67 6 3.0 50.0 2 2 2 33.3 33.3 3387
19 SFT 221006-CAB4_nnad9b 3388 0.1 60 70 6 3.5 58.3 2 1 3 33.3 50.0 3350
20 SF Official 221105-61a2ci_nna 3386 1.4 66 68 6 3.0 50.0 2 2 2 33.3 33.3 3387
21 Stockfish_230313_T902 nn5247 3383 3.0 58 70 6 3.5 58.3 2 1 3 33.3 50.0 3348
22 SFT 230302-c0a58-87690_nn5247 3382 1.2 63 64 6 3.0 50.0 1 1 4 16.7 66.7 3378
23 SFT 230306-2be47-cdec7_nn5247 3380 1.7 68 69 6 3.0 50.0 3 3 0 50.0 0.0 3379
24 Stockfish_230227_T830 nn1337 3380 0.6 66 66 6 3.0 50.0 1 1 4 16.7 66.7 3379
25 Stockfish_230227_T987 nn1337 3379 0.5 64 64 6 3.0 50.0 0 0 6 0.0 100.0 3379
26 SFofficial 221008-d527_nnad9b 3378 1.5 59 70 6 3.5 58.3 2 1 3 33.3 50.0 3342
27 Stockfish_221016_T901 nnad9b 3373 5.1 67 68 6 3.0 50.0 2 2 2 33.3 33.3 3372
28 Stockfish_220917_T335 nnad9b 3370 2.4 40 43 18 9.5 52.8 4 3 11 22.2 61.1 3356
29 Stockfish 170422 T285 nn6877 3370 0.3 24 25 42 25.5 60.7 15 6 21 35.7 50.0 3325
30 SFT 221005-ad76_nnad9b 3370 0.0 65 67 6 3.0 50.0 1 1 4 16.7 66.7 3370
31 SFT 230301-7ad71-87690_nn5247 3370 0.2 70 58 6 2.5 41.7 1 2 3 16.7 50.0 3408
32 Stockfish_22091112_T554 nnad9 3369 0.5 44 49 12 6.5 54.2 2 1 9 16.7 75.0 3350
33 Stockfish_221003_T447 nnad9b 3367 2.0 39 41 18 9.5 52.8 6 5 7 33.3 38.9 3354
34 SFT 220914-71df-82bb_nnad9b 3367 0.5 69 70 6 3.0 50.0 2 2 2 33.3 33.3 3367
35 SFT 220923-b0db5_nnad9b 3364 2.4 60 67 6 3.5 58.3 2 1 3 33.3 50.0 3333
36 Stockfish_221016_T928 nnad9b 3363 1.1 47 46 12 5.5 45.8 3 4 5 25.0 41.7 3378
37 SFofficial 221003-8bab_nnad9b 3363 0.6 40 40 18 9.0 50.0 4 4 10 22.2 55.6 3362
38 Stockfish_221016_T516 nnad9b 3362 0.4 66 58 6 2.5 41.7 0 1 5 0.0 83.3 3396
39 Stockfish_22090705_T708 nnad9 3362 0.4 36 37 22 11.0 50.0 4 4 14 18.2 63.6 3360
40 SFT 220401-CD140_nn6877 3361 0.7 24 25 46 24.0 52.2 9 7 30 19.6 65.2 3351
41 SF Offciail 220815-4568_nnad9 3361 0.6 32 32 30 15.0 50.0 8 8 14 26.7 46.7 3360
42 SFT 230303-80973-87690_nn5247 3359 1.8 66 66 6 3.0 50.0 1 1 4 16.7 66.7 3358
43 Stockfish_230305_T376 nn5247 3359 0.3 65 65 6 3.0 50.0 0 0 6 0.0 100.0 3359
44 SFT 221006-70C3_nnad9b 3357 1.8 59 65 6 3.5 58.3 1 0 5 16.7 83.3 3326
45 Stockfish_22050407_T276nnd0b7 3356 0.4 25 26 44 23.5 53.4 11 8 25 25.0 56.8 3340
46 Stockfish_220927_T685 nnad9b 3356 0.2 67 66 6 3.0 50.0 1 1 4 16.7 66.7 3356
47 SFT 220930-B76A-232BF nnad9b 3356 0.1 65 65 6 3.0 50.0 0 0 6 0.0 100.0 3356
48 Stockfish_22081518_T297 nnad9 3354 1.9 39 40 18 9.0 50.0 4 4 10 22.2 55.6 3353
49 SFT 230228-4fb74-87690_nn5247 3354 0.5 79 70 4 1.5 37.5 1 2 1 25.0 25.0 3399
50 Stockfish_22050317_T481nnd0b7 3353 0.6 26 27 40 20.5 51.2 6 5 29 15.0 72.5 3346

PGN DOWNLOAD LINK: https://www.file-upload.com/rw986ww92jh7
-----------------------------------------------

NEW ENGINED ADDED TO QUEUE

ENGINE NAME
SF Official 230401-6a6e3 master (Decrease Depth more for positions not in TT.)

ENGINE INFO
SF Official 230401-6a6e3 master, Decrease Depth more for positions not in TT. Elo: 52.68 +-2.6 (95%) over Stockfish_15. By the Stockfish developers (see AUTHORS file).

SPEED TEST (BMI2-PGO, INTEL I3 2.00 GHZ)
Total time (ms) : 20274
Nodes searched : 4380438
Nodes/second : 216061

DOWNLOAD LINK: https://www.file-upload.com/434tzjmsofgu

[Engine]

Thanks

[mig2004]

ENGINE ADDED TO QUEUE

SF Official 230401-e2be5-master( Update search.cpp beta2 beta3)

Update search.cpp
beta2 beta3
@FauziAkram
FauziAkram committed 2 days ago
1 parent 3f01e3f
commit e2be5d4

https://github.com/FauziAkram/Stockfish ... 0664070b71

ENGINE INFO
SF Official 230401-e2be5 e2be5-master( Update search.cpp beta2 beta3), by the Stockfish developers (see AUTHORS file)

SPEED TEST (BMI2-PGO INTEL I3 2 GHZ)
Total time (ms) : 19990
Nodes searched : 4646341
Nodes/second : 232433

download link: https://www.file-upload.com/9c0rzpfix4w6

[mig2004]

UPCOMING TOURNAMENT ENGINE

ORIGINAL INFO

23-04-03 fau tunee2a diff LLR: 2.94 (-2.94,2.94) <0.00,2.00>Total: 27056 W: 7415 L: 7125 D: 12516Ptnml(0-2): 76, 2835, 7420, 3117, 80 sprt @ 10+0.1 th 1 tunee2a 6a6e3-master (Decrease Depth more for positions not in TT.) 3.56 [1.49,5.63] (95%)

COMMIT INFO
Update search.cpp , tunee2a , FauziAkram committed Apr 3, 2023, 1 parent 6a6e32d commit 284360b
ENGINE RENAMED AS:
SFT 230403-28436-6a6e3-master (Decrease Depth more for positions not in TT.)

ENGINE INFO
SFT 230403-28436-6a6e3 Update search.cpp , tunee2a , FauziAkram committed Apr 3, 2023, 1 parent 6a6e32d commit 284360b, Elo gain: 3.56 [1.49,5.63] (95%) over master 6a6e3- (Decrease Depth more for positions not in TT.)the Stockfish developers (see AUTHORS file)

SPEED TEST (BMI2-PGO,INTEL I3, 2 GHZ)
Total time (ms) : 15688
Nodes searched : 3582149
Nodes/second : 228336

download link: https://www.file-upload.com/9ug5mvcsqgnc

[mig2004]

UPCOMING ENGINES

stockfish 230405-1fee9dev-Remove unneeded bitboard from MP.7z
481.5 MB
https://www.file-upload.com/qw32l335ndy6

stockfish_ 230405-77e2b-Simplifiy TM's root complexity.7z
481.5 MB
https://www.file-upload.com/c1mfiv3iunrn