Chess Forum

Author: Stéphane Nicolet
Date: Mon Sep 11 23:19:06 2023 +0200
Timestamp: 1694467146

Reformat some comments

Also include the bench to make Continuation Integration happy on Github.

Bench: 1603079

Author: Linmiao Xu
Date: Fri Sep 22 19:26:16 2023 +0200
Timestamp: 1695403576

Update NNUE architecture to SFNNv8: L1-2560 nn-ac1dbea57aa3.nnue

Creating this net involved:
- a 6-stage training process from scratch. The datasets used in stages 1-5 were fully minimized.
- permuting L1 weights with https://github.com/official-stockfish/n ... h/pull/254

A strong epoch after each training stage was chosen for the next. The 6 stages were:

```
1. 400 epochs, lambda 1.0, default LR and gamma
UHOx2-wIsRight-multinet-dfrc-n5000 (135G)
nodes5000pv2_UHO.binpack
data_pv-2_diff-100_nodes-5000.binpack
wrongIsRight_nodes5000pv2.binpack
multinet_pv-2_diff-100_nodes-5000.binpack
dfrc_n5000.binpack

2. 800 epochs, end-lambda 0.75, LR 4.375e-4, gamma 0.995, skip 12
LeelaFarseer-T78juntoaugT79marT80dec.binpack (141G)
T60T70wIsRightFarseerT60T74T75T76.binpack
test78-junjulaug2022-16tb7p.no-db.min.binpack
test79-mar2022-16tb7p.no-db.min.binpack
test80-dec2022-16tb7p.no-db.min.binpack

3. 800 epochs, end-lambda 0.725, LR 4.375e-4, gamma 0.995, skip 20
leela93-v1-dfrc99-v2-T78juntosepT80jan-v6dd-T78janfebT79aprT80aprmay.min.binpack
leela93-filt-v1.min.binpack
dfrc99-16tb7p-filt-v2.min.binpack
test78-juntosep2022-16tb7p-filter-v6-dd.min-mar2023.binpack
test80-jan2023-3of3-16tb7p-filter-v6-dd.min-mar2023.binpack
test78-janfeb2022-16tb7p.min.binpack
test79-apr2022-16tb7p.min.binpack
test80-apr2022-16tb7p.min.binpack
test80-may2022-16tb7p.min.binpack

4. 800 epochs, end-lambda 0.7, LR 4.375e-4, gamma 0.995, skip 24
leela96-dfrc99-v2-T78juntosepT79mayT80junsepnovjan-v6dd-T80mar23-v6-T60novdecT77decT78aprmayT79aprT80may23.min.binpack
leela96-filt-v2.min.binpack
dfrc99-16tb7p-filt-v2.min.binpack
test78-juntosep2022-16tb7p-filter-v6-dd.min-mar2023.binpack
test79-may2022-16tb7p.filter-v6-dd.min.binpack
test80-jun2022-16tb7p.filter-v6-dd.min.binpack
test80-sep2022-16tb7p.filter-v6-dd.min.binpack
test80-nov2022-16tb7p.filter-v6-dd.min.binpack
test80-jan2023-3of3-16tb7p-filter-v6-dd.min-mar2023.binpack
test80-mar2023-2tb7p.v6-sk16.min.binpack
test60-novdec2021-16tb7p.min.binpack
test77-dec2021-16tb7p.min.binpack
test78-aprmay2022-16tb7p.min.binpack
test79-apr2022-16tb7p.min.binpack
test80-may2023-2tb7p.min.binpack

5. 960 epochs, end-lambda 0.7, LR 4.375e-4, gamma 0.995, skip 28
Increased max-epoch to 960 near the end of the first 800 epochs
5af11540bbfe dataset: https://github.com/official-stockfish/S ... /pull/4635

6. 1000 epochs, end-lambda 0.7, LR 4.375e-4, gamma 0.995, skip 28
Increased max-epoch to 1000 near the end of the first 800 epochs
1ee1aba5ed dataset: https://github.com/official-stockfish/S ... /pull/4782
```

L1 weights permuted with:
```bash
python3 serialize.py $nnue $nnue_permuted \
--features=HalfKAv2_hm \
--ft_optimize \
--ft_optimize_data=/data/fishpack32.binpack \
--ft_optimize_count=10000
```

Speed measurements from 100 bench runs at depth 13 with profile-build x86-64-avx2:
```
sf_base = 1329051 +/- 2224 (95%)
sf_test = 1163344 +/- 2992 (95%)
diff = -165706 +/- 4913 (95%)
speedup = -12.46807% +/- 0.370% (95%)
```

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

Local elo at 25k nodes per move (vs. L1-2048 nn-1ee1aba5ed4c.nnue)
ep959 : 16.2 +/- 2.3

Failed 10+0.1 STC:
https://tests.stockfishchess.org/tests/ ... da89abab21
LLR: -2.92 (-2.94,2.94) <0.00,2.00>
Total: 13184 W: 3285 L: 3535 D: 6364 Elo -6.59
Ptnml(0-2): 85, 1662, 3334, 1440, 71

Failed 180+1.8 VLTC:
https://tests.stockfishchess.org/tests/ ... c881ea908e
LLR: -2.94 (-2.94,2.94) <0.00,2.00>
Total: 64248 W: 16224 L: 16374 D: 31650 Elo -0.81
Ptnml(0-2): 26, 6788, 18640, 6650, 20

Passed 60+0.6 th 8 VLTC SMP (STC bounds):
https://tests.stockfishchess.org/tests/ ... 74c2e541dc
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 90630 W: 23372 L: 23033 D: 44225 Elo +1.30
Ptnml(0-2): 13, 8490, 27968, 8833, 11

Passed 60+0.6 th 8 VLTC SMP:
https://tests.stockfishchess.org/tests/ ... da89abacdb
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 137804 W: 35764 L: 35276 D: 66764 Elo +1.23
Ptnml(0-2): 31, 13006, 42326, 13522, 17

closes https://github.com/official-stockfish/S ... /pull/4795

bench 1246812

Author: Joost VandeVondele
Date: Sun Sep 24 20:04:42 2023 +0200
Timestamp: 1695578682

Explicitly invoke shell

in some cases the permission on the script might be incorrect (zip downloads?).
Explicitly invoke the shell

closes https://github.com/official-stockfish/S ... /pull/4803

No functional change

Author: Linmiao Xu
Date: Fri Sep 29 22:30:27 2023 +0200
Timestamp: 1696019427

Update default net to nn-0000000000a0.nnue

This is a later epoch from the same experiment that led to the previous
master net. In training stage 6, max-epoch was raised to 1,200 near the
end of the first 1,000 epochs.

For more details, see https://github.com/official-stockfish/S ... /pull/4795

Local elo at 25k nodes per move (vs. L1-2048 nn-1ee1aba5ed4c.nnue)
ep1079 : 15.6 +/- 1.2

Passed STC:
https://tests.stockfishchess.org/tests/ ... 11c8af1e2a
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 29408 W: 7607 L: 7304 D: 14497 Elo +3.58
Ptnml(0-2): 97, 3277, 7650, 3586, 94

Passed LTC:
https://tests.stockfishchess.org/tests/ ... 11c8af2a5f
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 73164 W: 18828 L: 18440 D: 35896 Elo +1.84
Ptnml(0-2): 30, 7749, 20644, 8121, 38

closes https://github.com/official-stockfish/S ... /pull/4810

Bench: 1453057

Author: Taras Vuk
Date: Sun Oct 8 07:56:07 2023 +0200
Timestamp: 1696744567

Skip futility pruning if ttMove has bad history

Passed STC:
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 52416 W: 13465 L: 13128 D: 25823 Elo +2.23
Ptnml(0-2): 128, 6024, 13604, 6287, 165
https://tests.stockfishchess.org/tests/ ... 14367277bf

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 87348 W: 22234 L: 21818 D: 43296 Elo +1.65
Ptnml(0-2): 38, 9240, 24698, 9664, 34
https://tests.stockfishchess.org/tests/ ... 1436728218

closes https://github.com/official-stockfish/S ... /pull/4825

bench: 1246560

Author: Muzhen Gaming
Date: Tue Oct 10 17:45:32 2023 +0200
Timestamp: 1696952732

Search parameters tune at 180+1.8

Passed VLTC: https://tests.stockfishchess.org/tests/ ... 14367280bc
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 146180 W: 37407 L: 36988 D: 71785 Elo +1.00
Ptnml(0-2): 21, 14474, 43675, 14905, 15

Passed VLTC SMP: https://tests.stockfishchess.org/tests/ ... 8fc7eb4b6d
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 57580 W: 15061 L: 14739 D: 27780 Elo +1.94
Ptnml(0-2): 2, 5001, 18460, 5327, 0

closes https://github.com/official-stockfish/S ... /pull/4826

Bench: 1099336

Author: Michael Chaly
Date: Sat Oct 14 16:52:35 2023 +0200
Timestamp: 1697295155

Use more continuation histories.

This patch allows stats updates and movepicker bonuses for continuation history 3 plies deep - so counter counter move.
Updates and movepicker usage are done with 1/4 multiplier compared to other histories.

Passed STC:
https://tests.stockfishchess.org/tests/ ... 8fc7ebb5a3
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 161344 W: 41369 L: 40868 D: 79107 Elo +1.08
Ptnml(0-2): 535, 18720, 41679, 19185, 553

Passed LTC:
https://tests.stockfishchess.org/tests/ ... 8fc7ebd1d6
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 48564 W: 12556 L: 12215 D: 23793 Elo +2.44
Ptnml(0-2): 25, 5149, 13595, 5486, 27

closes https://github.com/official-stockfish/S ... /pull/4827

bench 1327410

massimilianogoi wrote: ↑Sat Oct 14, 2023 5:55 pm New rule to respect in this single topic: any post must include a link to download the new compilation of rhw program (like a link the the Github releases). This to prevent the continue flooding of the board.

With some engines/databases its a good idea to have links to NOT JUST the executable - there may be important considerations/information that I personally like to be made aware of.

Case in point .. this is important and would have been missed if ONLY the executable was linked to .. in my opinion of course.

From vondele >>

The right thing here is to setoption name Move Overhead value 1500, i.e. the option is exactly there to inform about expected (worst case?) GUI delays.

Receiving negative (even including zero) btime could lead to all kind of weird stuff. This is really a GUI bug.

That's right, but it's annoying to see this topic flooded with offtopic stuff. It is titled "Stockfish download" for some reason.

I could make an exception if the changing/bug is relevant, otherwise the link to the new compiled binaries is mandatory. Besided they have the assets page on github, where all these compiled executables are available.

For example: https://github.com/official-stockfish/S ... ase%3Atrue .

janus wrote: ↑Sat Oct 14, 2023 8:21 pm
With some engines/databases its a good idea to have links to NOT JUST the executable - there may be important considerations/information that I personally like to be made aware of.

Case in point .. this is important and would have been missed if ONLY the executable was linked to .. in my opinion of course.

From vondele >>

The right thing here is to setoption name Move Overhead value 1500, i.e. the option is exactly there to inform about expected (worst case?) GUI delays.

Receiving negative (even including zero) btime could lead to all kind of weird stuff. This is really a GUI bug.