edbbc1a4df
Remove some warnings
2020-09-07 09:20:47 +09:00
3bf418e63f
Fix some uninitialized variables with gensfen
...
fixes valgrind errors as seen with:
```
setoption name Use NNUE value true
isready
gensfen depth 6 loop 10 use_draw_in_training_data_generation 1 eval_limit 32000 output_file_name training_data/training_data.bin use_raw_nnue_eval 0
quit
```
the latter script now runs without valgrind errors on linux
2020-09-07 09:01:17 +09:00
e9e6e47a93
Fix write_out_draw_game_in_training_data_generation flag not being respected.
2020-09-06 22:00:51 +09:00
0612adec41
Fix incorrect early exit in evaluate_leaf.
2020-09-05 08:43:34 +09:00
327e92aefe
Remove trailing whitespaces.
2020-09-03 19:22:52 +09:00
2688194d44
Fix #91
2020-09-03 19:22:03 +09:00
9d5dc3d33f
Fix compilation issues.
2020-09-03 19:21:27 +09:00
c17f2b15fd
General cleanup of learner.cpp.
2020-09-02 23:08:22 +09:00
f7bc4e6e45
Fixed compilation errors.
2020-08-29 00:56:05 +09:00
5637996f79
Resolve #92
...
If we're defining something in a header then we should declare it.
2020-08-28 11:20:03 +09:00
03b43079eb
Merge branch 'nodchip_master' into wdl_20200824
2020-08-28 09:16:07 +09:00
763e72cc9f
Changed the default value of use_raw_nnue_eval. Updated a source code comment.
2020-08-27 23:49:58 +09:00
9fc3ff4c30
Added use_raw_nnue_eval option to return raw NNUE eval value in evaluate().
2020-08-27 23:48:28 +09:00
4ce30d9522
Use winning_percentage_wdl in learn
2020-08-24 22:56:08 +09:00
7ee8a2bbb7
Merge branch 'master' of github.com:nodchip/Stockfish
2020-08-22 12:19:46 +09:00
87633b876c
Added an option to convert teacher signals to winning probabilities.
2020-08-22 12:19:29 +09:00
f057aec4a9
Merge remote-tracking branch 'nodchip/master' into convert_bin_from_pgn-extract_20200819
2020-08-21 22:49:11 +09:00
11752d4e63
Added options to scale the scores in training data. #71
2020-08-21 21:16:55 +09:00
aaa73b2569
modify convert_bin_from_pgn-extract
2020-08-19 22:47:22 +09:00
2646080543
Merge branch 'master' of github.com:nodchip/Stockfish
2020-08-15 11:58:05 +09:00
79654ac509
Added winning_probability_coefficient option to specify the coefficient to calculate the winning probability from a value. #71
2020-08-15 11:57:08 +09:00
69a95e431b
update convert_bin
...
learn convert_bin in.txt output_file_name out.bin check_invalid_fen 1 check_illegal_move 1
convert in.txt ... done 16 parsed 3 is filtered (invalid fen:1, illegal move:2, invalid ply:0)
2020-08-15 00:27:52 +09:00
35f04aaf24
Removed an unnecessary call for pos.is_draw().
2020-08-10 19:42:39 +09:00
c420b327bf
Added output messages.
2020-08-10 16:23:04 +09:00
8c0429d1e5
Added detect_draw_by_insufficient_mating_material option.
2020-08-10 16:14:56 +09:00
84070c02e6
Renamed use_game_draw_adjudication to detect_draw_by_consecutive_low_score.
2020-08-10 16:02:18 +09:00
5467ba3c23
Renamed use_hash_in_training to skip_duplicated_positions_in_training.
2020-08-10 15:58:17 +09:00
87c50c5cbc
Renamed use_draw_in_validation to use_draw_games_in_validation.
...
Added comments.
2020-08-10 15:55:34 +09:00
fa5b2aec3a
Renamed use_draw_in_training to use_draw_games_in_training.
2020-08-10 15:51:23 +09:00
a41cbb9ca9
Renamed use_draw_in_training_data_generation option to write_out_draw_game_in_training_data_generation.
2020-08-10 15:49:24 +09:00
3bd3ef0aea
Implemented the code to detect draw by insufficient mating material.
2020-08-10 15:47:11 +09:00
4a87d7b787
Added the use_game_draw_adjudication option.
2020-08-10 15:44:58 +09:00
12c6c2f550
Chagned to use the search value instead of the value of the PV leaf to avoid crash by assertion.
2020-08-10 13:07:22 +09:00
bac96aa04a
Changed to use TB in the training data generator. #67
2020-08-10 12:17:26 +09:00
e65c515d6b
Changed to specify the current tick as a random seed. #68
2020-08-10 12:09:21 +09:00
643be3c6f9
Changed not to use std::random_device(). Because it always returns the same integers on MingW. #68
2020-08-10 10:45:03 +09:00
31d4f46f5e
update convert_bin
...
learn convert_bin in.txt output_file_name out.bin check_illegal_move 1
convert in.txt ... done 16 parsed 4 is filtered (illegal fen:1, illegal move:2, illegal ply:1)
2020-08-10 09:53:52 +09:00
4f97d3446d
Cleaned up source code.
2020-08-10 08:52:34 +09:00
4f94f29f39
Revert "Fixed a bug that the training data generation crashes if eval_limit is high."
...
This reverts commit b0d28ac3ab
2020-08-08 18:38:02 +09:00
9a0b20d3fc
Changed to show if NNUE is used in the training data generator.
2020-08-08 18:24:09 +09:00
b0d28ac3ab
Fixed a bug that the training data generation crashes if eval_limit is high.
2020-08-08 18:23:11 +09:00
70d88364fe
Fixed a bug that the training data generation crashes.
2020-08-08 18:22:29 +09:00
55a6b2bdc4
Merge branch 'master' of github.com:official-stockfish/Stockfish into nnue-player-merge
...
# Conflicts:
# README.md
# Readme.md
# src/Makefile
# src/evaluate.cpp
# src/evaluate.h
# src/misc.cpp
# src/nnue/architectures/halfkp_256x2-32-32.h
# src/nnue/evaluate_nnue.cpp
# src/nnue/evaluate_nnue.h
# src/nnue/features/feature_set.h
# src/nnue/features/features_common.h
# src/nnue/features/half_kp.cpp
# src/nnue/features/half_kp.h
# src/nnue/features/index_list.h
# src/nnue/layers/affine_transform.h
# src/nnue/layers/clipped_relu.h
# src/nnue/layers/input_slice.h
# src/nnue/nnue_accumulator.h
# src/nnue/nnue_architecture.h
# src/nnue/nnue_common.h
# src/nnue/nnue_feature_transformer.h
# src/position.cpp
# src/position.h
# src/types.h
# src/ucioption.cpp
# stockfish.md
2020-08-08 15:55:42 +09:00
bf7d02578e
Fixed build errors.
2020-08-07 22:47:45 +09:00
686a5a0df9
Fixed a bug that gensfen command does not accept the use_draw_in_training_data_generation option.
2020-07-13 22:25:23 +09:00
a06234c639
enable convert_plain
...
learn convert_plain output_file_name xxx.txt xxx.bin
2020-07-09 09:51:00 +09:00
a5af8510a5
Rework loading the net.
2020-07-07 15:13:59 +09:00
288fdc5597
Added "nodes" option to the "gensfen" command to specify the number of the nodes to be searched.
2020-07-06 17:38:43 +09:00
85c802d0b9
Revert "use winning_percentage_wdl in learn"
...
This reverts commit c964e902c5
2020-07-06 11:07:46 +09:00
c964e902c5
use winning_percentage_wdl in learn
2020-07-03 23:21:49 +09:00
Joost VandeVondele