Skip to content

guitanj/go-test

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

46 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Usage

Change the command in config.yml to use your own Go program command (Only leela zero supported now), you can also use --command option to override the config.

./run.py will run all tests

./run.py -h for help

Example output:

Boods-MacBook-Pro:go-test Bood$ ./run.py
Command: /Users/Bood/mypro/leela-zero/bin/leelaz -d -t 1 -p 1600 --noponder -w /Users/Bood/mypro/leela-zero/bin/af9
ladder1.sgf
[FAIL]  C17 ->     254 (V: 39.95%) (N:  8.68%) PV: C17 B15 E18 D17 D18 B17 B18 C16 C18 F18 A17 B16 P7 O7
ladder2.sgf
[PASS]   E1 ->     893 (V: 40.28%) (N: 42.68%) PV: E1 G4 E9 E10 F10 E11 D9 C12 D14 F9 C9
ladder3.sgf
[FAIL]  P11 ->     708 (V: 57.36%) (N:  0.17%) PV: P11 N11 O12 O13 N12 M12 N13 N14 M13 L13 M14 M15 L14 K14 L15 L16 K15
ladder4.sgf
[PASS]  C16 ->     660 (V: 27.50%) (N:  0.09%) PV: C16 P6 D17 P2 Q2 F3 P3 O3
ladder5.sgf
[PASS]   O7 ->     792 (V: 45.98%) (N: 23.54%) PV: O7 J6 P6 O5 C11 C9 E12 E13 F13
lifedeath1.sgf
[FAIL]   M9 ->    1318 (V: 97.09%) (N: 61.30%) PV: M9 S18 T18 S19 R19 S16 T19 S15 S14 R14 T15 Q16 T16 Q15
longdragon1.sgf
[FAIL]   D8 ->     524 (V: 96.66%) (N:  6.78%) PV: D8 D9 D7 C9 E9 D10 F9 G9
longdragon2.sgf
[FAIL]   N5 ->     754 (V: 94.61%) (N:  0.40%) PV: N5 L5 O6 T2 L17 J5 K18 K16

Add tests

  1. Put your sgf file in sgf dir, e.g. ladder6.sgf

  2. Update config.yml with an entry in tests section, with following information:

    • name test name
    • sgf file name
    • group test group it belongs to
    • number which move number to start with
    • move who's turn it is now
    • test condition, one of the following (only one will be used):
      • yes_move: a list of moves it should choose, fail otherwise
      • no_move: a list of moves it should NOT choose, fail if the choosed move is in the list
      • max_win_rate: upper threshold of win rate of value network, fail if the program thinks it's higher
  3. Verify your added test with ./run.py --case test_name

About

Scenario based strength test tool for Go programs

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 100.0%