mirror of
https://github.com/sockspls/badfish
synced 2025-05-01 01:03:09 +00:00

Created by training L1-128 from scratch with: - skipping based on simple eval in the trainer, for compatibility with regular binpacks without requiring pre-filtering all binpacks - minimum simple eval of 950, lower than 1000 previously - usage of some hse-v1 binpacks with minimum simple eval 1000 - addition of hse-v6 binpacks with minimum simple eval 500 - permuting the FT with 10k positions from fishpack32.binpack - torch.compile to speed up smallnet training Training is significantly slower when using non-pre-filtered binpacks due to the increased skipping required. This net was reached at epoch 339. ``` experiment-name: 128--S1-hse-1k-T80-v6-unfilt-less-sf--se-gt950-no-wld-skip training-dataset: /data/: - dfrc99-16tb7p.v2.min.binpack /data/hse-v1/: - leela96-filt-v2.min.high-simple-eval-1k.min-v2.binpack - test60-novdec2021-12tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test77-nov2021-2tb7p.no-db.min.high-simple-eval-1k.min-v2.binpack - test77-dec2021-16tb7p.no-db.min.high-simple-eval-1k.min-v2.binpack - test77-jan2022-2tb7p.high-simple-eval-1k.min-v2.binpack - test78-jantomay2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test78-juntosep2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test79-apr2022-16tb7p.min.high-simple-eval-1k.min-v2.binpack - test79-may2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test80-apr2022-16tb7p.min.high-simple-eval-1k.min-v2.binpack - test80-may2022-16tb7p.high-simple-eval-1k.min-v2.binpack - test80-jun2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test80-jul2022-16tb7p.v6-dd.min.high-simple-eval-1k.min-v2.binpack - test80-sep2022-16tb7p-filter-v6-dd.min-mar2023.unmin.high-simple-eval-1k.min-v2.binpack - test80-nov2022-16tb7p-v6-dd.min.high-simple-eval-1k.min-v2.binpack /data/S11-mar2024/: - test80-2022-08-aug-16tb7p.v6-dd.min.binpack - test80-2022-10-oct-16tb7p.v6-dd.binpack - test80-2022-12-dec-16tb7p.min.binpack - test80-2023-01-jan-16tb7p.v6-sk20.min.binpack - test80-2023-02-feb-16tb7p.v6-sk20.min.binpack - test80-2023-03-mar-2tb7p.v6-sk16.min.binpack - test80-2023-04-apr-2tb7p.v6-sk16.min.binpack - test80-2023-05-may-2tb7p.v6.min.binpack - test80-2023-06-jun-2tb7p.binpack.min-v2.binpack - test80-2023-07-jul-2tb7p.binpack.min-v2.binpack - test80-2023-08-aug-2tb7p.v6.min.binpack - test80-2023-09-sep-2tb7p.binpack.hse-v6.binpack - test80-2023-10-oct-2tb7p.binpack.hse-v6.binpack - test80-2023-11-nov-2tb7p.binpack.hse-v6.binpack - test80-2023-12-dec-2tb7p.binpack.hse-v6.binpack - test80-2024-01-jan-2tb7p.binpack.hse-v6.binpack - test80-2024-02-feb-2tb7p.binpack.hse-v6.binpack - test80-2024-03-mar-2tb7p.binpack wld-fen-skipping: False nnue-pytorch-branch: linrock/nnue-pytorch/128-skipSimpleEval-lt950-torch-compile engine-test-branch: linrock/Stockfish/L1-128-nolazy engine-base-branch: linrock/Stockfish/L1-128 start-from-engine-test-net: False num-epochs: 500 start-lambda: 1.0 end-lambda: 1.0 ``` Training data can be found at: https://robotmoon.com/nnue-training-data/ Passed STC: https://tests.stockfishchess.org/tests/view/66549c16a86388d5e27daff5 LLR: 2.93 (-2.94,2.94) <0.00,2.00> Total: 196608 W: 51254 L: 50697 D: 94657 Ptnml(0-2): 722, 23244, 49796, 23839, 703 Passed LTC: https://tests.stockfishchess.org/tests/view/6658d1aa6b0e318cefa90122 LLR: 2.96 (-2.94,2.94) <0.50,2.50> Total: 122538 W: 31332 L: 30835 D: 60371 Ptnml(0-2): 69, 13407, 33811, 13922, 60 closes https://github.com/official-stockfish/Stockfish/pull/5333 bench
56 lines
1.8 KiB
C++
56 lines
1.8 KiB
C++
/*
|
|
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
|
|
Copyright (C) 2004-2024 The Stockfish developers (see AUTHORS file)
|
|
|
|
Stockfish is free software: you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
(at your option) any later version.
|
|
|
|
Stockfish is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
|
|
#ifndef EVALUATE_H_INCLUDED
|
|
#define EVALUATE_H_INCLUDED
|
|
|
|
#include <string>
|
|
|
|
#include "types.h"
|
|
|
|
namespace Stockfish {
|
|
|
|
class Position;
|
|
|
|
namespace Eval {
|
|
|
|
// The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
|
|
// for the build process (profile-build and fishtest) to work. Do not change the
|
|
// name of the macro or the location where this macro is defined, as it is used
|
|
// in the Makefile/Fishtest.
|
|
#define EvalFileDefaultNameBig "nn-ddcfb9224cdb.nnue"
|
|
#define EvalFileDefaultNameSmall "nn-37f18f62d772.nnue"
|
|
|
|
namespace NNUE {
|
|
struct Networks;
|
|
struct AccumulatorCaches;
|
|
}
|
|
|
|
std::string trace(Position& pos, const Eval::NNUE::Networks& networks);
|
|
|
|
int simple_eval(const Position& pos, Color c);
|
|
bool use_smallnet(const Position& pos);
|
|
Value evaluate(const NNUE::Networks& networks,
|
|
const Position& pos,
|
|
Eval::NNUE::AccumulatorCaches& caches,
|
|
int optimism);
|
|
} // namespace Eval
|
|
|
|
} // namespace Stockfish
|
|
|
|
#endif // #ifndef EVALUATE_H_INCLUDED
|