Since you’re here...

We hope you will consider supporting us today. We need your support to continue to exist, because good entries are more and more work time. Every reader contribution, however big or small, is so valuable. Support "Chess Engines Diary" even a small amount– and it only takes a minute. Thank you.
============================== My email: jotes@go2.pl



Xiangqi engine: Pikafish 2024-03-10 (Windows, MacOS, Linux and Android)

on


Pikafish is a free, powerful UCI xiangqi engine derived from Stockfish. Pikafish is not a complete xiangqi program and requires a UCI-compatible graphical user interface (GUI) (e.g. VinXiangQi or BHGUI) in order to be used comfortably. Read the documentation for your GUI of choice for information about how to use Pikafish with it.

The Pikafish engine features the efficiently updatable neural network (NNUE) based evaluation for xiangqi, which is by far the strongest. The strongest network can be downloaded from our official website. The NNUE evaluation benefits from the vector intrinsics available on most CPUs (sse2, avx2, neon, or similar).

v.2024.03.10:
Comparing with Pikafish 2023-12-03:
Book: 3mvs 151-350_1097123

TC: 10+0.1
Total/Win/Draw/Lose: 4036 / 1165 / 1979 / 892
PTNML: 33 / 382 / 942 / 601 / 60
WinRate: 53.38%
ELO: 23.28[16.64, 30.04]
LOS: 100.00
LLR: 3.35[-2.94, 2.94]

TC: 60+0.6
Total/Win/Draw/Lose: 2124 / 597 / 1133 / 394
PTNML: 2 / 162 / 544 / 339 / 15
WinRate: 54.78%
ELO: 33.06[25.41, 41.01]
LOS: 100.00
LLR: 3.02[-2.94, 2.94]

Book: 3mvs 151-350_1097123
TC: 180+1.8
Total/Win/Draw/Lose: 1414 / 432 / 706 / 276
PTNML: 0 / 76 / 402 / 226 / 3
WinRate: 55.52%
ELO: 38.23[29.85, 47.12]
LOS: 100.00
LLR: 2.97[-2.94, 2.94]

Shows very good scaling.

Pikafish 2024-03-10 download


Comments

Post a Comment