kate/BadFish
1
0
Fork 0
mirror of https://github.com/sockspls/badfish synced 2025-04-29 16:23:09 +00:00
Code Issues Releases Wiki Activity

Lower smallnet threshold with tuned eval params

Browse source 
The smallnet threshold is now below the training data range
of the current smallnet (simple eval diff > 1k, nn-baff1edelf90.nnue)
when no pawns are on the board.

Params found with spsa at 93k / 120k games at 60+06:
https://tests.stockfishchess.org/tests/view/664fa166a86388d5e27d7d6b

Tuned on top of: https://github.com/official-stockfish/Stockfish/pull/5287

Passed STC:
https://tests.stockfishchess.org/tests/view/664fc8b7a86388d5e27d8dac
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 64672 W: 16731 L: 16371 D: 31570
Ptnml(0-2): 239, 7463, 16517, 7933, 184

Passed LTC:
https://tests.stockfishchess.org/tests/view/664fd5f9a86388d5e27d8dfe
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 210648 W: 53489 L: 52813 D: 104346
Ptnml(0-2): 102, 23129, 58164, 23849, 80

closes https://github.com/official-stockfish/Stockfish/pull/5288

Bench: 1717838
This commit is contained in:
Linmiao Xu 2024-05-24 10:58:13 -04:00 • committed by Disservin
parent 4d876275cf
commit 8bc3fd3871
1 changed files with 6 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
src/evaluate.cpp
View file

@ -46,7 +46,7 @@ int Eval::simple_eval(const Position& pos, Color c) {
bool Eval::use_smallnet(const Position& pos) { bool Eval::use_smallnet(const Position& pos) {
int simpleEval = simple_eval(pos, pos.side_to_move()); int simpleEval = simple_eval(pos, pos.side_to_move());
return std::abs(simpleEval) > 1018 + 5 * pos.count<PAWN>(); return std::abs(simpleEval) > 992 + 6 * pos.count<PAWN>();
} }
// Evaluate is the evaluator for the outer world. It returns a static evaluation // Evaluate is the evaluator for the outer world. It returns a static evaluation
@ -74,13 +74,15 @@ Value Eval::evaluate(const Eval::NNUE::Networks& networks,
} }
// Blend optimism and eval with nnue complexity // Blend optimism and eval with nnue complexity
optimism += optimism * nnueComplexity / 512; optimism += optimism * nnueComplexity / 470;
nnue -= nnue * (nnueComplexity * 5 / 3) / 32082; nnue -= nnue * (nnueComplexity * 5 / 3) / 32621;
int material = 200 * pos.count<PAWN>() + 350 * pos.count<KNIGHT>() + 400 * pos.count<BISHOP>() int material = 200 * pos.count<PAWN>() + 350 * pos.count<KNIGHT>() + 400 * pos.count<BISHOP>()
+ 640 * pos.count<ROOK>() + 1200 * pos.count<QUEEN>(); + 640 * pos.count<ROOK>() + 1200 * pos.count<QUEEN>();
v = (nnue * (34000 + material) + optimism * (4400 + material)) / 36860; v = (nnue * (34000 + material + 135 * pos.count<PAWN>())
+ optimism * (4400 + material + 99 * pos.count<PAWN>()))
/ 35967;
// Damp down the evaluation linearly when shuffling // Damp down the evaluation linearly when shuffling
v = v * (204 - pos.rule50_count()) / 208; v = v * (204 - pos.rule50_count()) / 208;