Cutechess-cli Syzygy tablebases

[maximmasiutin]

The new worker option --syzygy_path to supply cutechess-cli (but not the engines) with the Syzygy tablebases path to use for adjudication to finish the games faster.

I tested this option and the majority of games are now finished by cutechess-cli with Syzygy adjudication, so it allows for the games to be resolved faster. This is a reasonable approach to supply the application that runs the engines but not the engines themselves with the tablebases. For example, it is used by TCEC.

Here is a snapshot of my output of the Fishtest worker showing the majority of the games are adjudicated using Syzygy tablebases before they could have finished by themselves after a certain additional number of moves:

Score of New-d92eeb485c vs Base-acb0d204d5: 3 - 1 - 5  [0.611] 9
Started game 11 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 10 (Base-acb0d204d5 vs New-d92eeb485c): 1/2-1/2 {Draw by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 3 - 1 - 6  [0.600] 10
Started game 12 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 11 (New-d92eeb485c vs Base-acb0d204d5): 1-0 {White wins by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 4 - 1 - 6  [0.636] 11
Started game 13 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 12 (Base-acb0d204d5 vs New-d92eeb485c): 1/2-1/2 {Draw by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 4 - 1 - 7  [0.625] 12
Started game 14 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 13 (New-d92eeb485c vs Base-acb0d204d5): 1/2-1/2 {Draw by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 4 - 1 - 8  [0.615] 13
Started game 15 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 14 (Base-acb0d204d5 vs New-d92eeb485c): 0-1 {White loses on time}
Score of New-d92eeb485c vs Base-acb0d204d5: 5 - 1 - 8  [0.643] 14
Started game 16 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 15 (New-d92eeb485c vs Base-acb0d204d5): 1/2-1/2 {Draw by 3-fold repetition}
Score of New-d92eeb485c vs Base-acb0d204d5: 5 - 1 - 9  [0.633] 15
Started game 17 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 16 (Base-acb0d204d5 vs New-d92eeb485c): 1-0 {White wins by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 5 - 2 - 9  [0.594] 16
Started game 18 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 17 (New-d92eeb485c vs Base-acb0d204d5): 1-0 {White wins by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 6 - 2 - 9  [0.618] 17
Started game 19 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 18 (Base-acb0d204d5 vs New-d92eeb485c): 1/2-1/2 {Draw by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 6 - 2 - 10  [0.611] 18
Started game 20 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 19 (New-d92eeb485c vs Base-acb0d204d5): 1/2-1/2 {Draw by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 6 - 2 - 11  [0.605] 19
Started game 21 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 20 (Base-acb0d204d5 vs New-d92eeb485c): 1-0 {White wins by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 6 - 3 - 11  [0.575] 20
Started game 22 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 21 (New-d92eeb485c vs Base-acb0d204d5): 1-0 {White wins by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 7 - 3 - 11  [0.595] 21
Started game 23 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 23 (New-d92eeb485c vs Base-acb0d204d5): 1/2-1/2 {Draw by 3-fold repetition}
Score of New-d92eeb485c vs Base-acb0d204d5: 7 - 3 - 12  [0.591] 22
Started game 24 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 22 (Base-acb0d204d5 vs New-d92eeb485c): 1-0 {White wins by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 7 - 4 - 12  [0.565] 23
Started game 25 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 24 (Base-acb0d204d5 vs New-d92eeb485c): 1-0 {White wins by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 7 - 5 - 12  [0.542] 24
Started game 26 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 25 (New-d92eeb485c vs Base-acb0d204d5): 1-0 {White wins by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 8 - 5 - 12  [0.560] 25
Started game 27 of 120 (New-d92eeb485c vs Base-acb0d204d5)
Finished game 26 (Base-acb0d204d5 vs New-d92eeb485c): 1/2-1/2 {Draw by adjudication}
Score of New-d92eeb485c vs Base-acb0d204d5: 8 - 5 - 13  [0.558] 26
Started game 28 of 120 (Base-acb0d204d5 vs New-d92eeb485c)
Finished game 27 (New-d92eeb485c vs Base-acb0d204d5): 1/2-1/2 {Draw by adjudication: SyzygyTB}
Score of New-d92eeb485c vs Base-acb0d204d5: 8 - 5 - 14  [0.556] 27
Started game 29 of 120 (New-d92eeb485c vs Base-acb0d204d5)

[maximmasiutin]

I tried this pull files request with pylint and all the files are OK. I didn't update the sri.txt file to avoid the conflicts. However, the CI pipelines seem to not like it. Can we somehow configure the pipelines to handle this issue, so that only the maintainer will update the file after merging probably several pull requests?