Run pip install -r requirements.txt to install all code requirements.
As an alternative to a web broswer, you may install Gaplin (http://gapplin.wolfrosch.com/) to open svgs.
Usage: python main.py -w <white/book/tbs> -b <black/book/tbs> -s <svgfile> -f <fen>
Example: python main.py -w hp -b mp/F/F -s game.svg -f '8/8/8/4p3/8/4k3/8/4K3 b - - 0 1'
| Option | Description | Default |
|---|---|---|
<white> |
the player type of White | mp (Minimax) |
<black> |
the player type of Black | gp (Greedy) |
<book> |
opening book usage (T/F) | T (True) |
<tbs> |
endgame tablebase usage (T/F) | T (True) |
<svgfile> |
outfile for svgs | "" (no svg) |
<fen> |
FEN for custom board position | chess.STARTING_FEN |
Player types:
Human (hp), Random (rp), Greedy (gp), Minimax (mp), ANN (nn)
Notes:
- If you don't use the
<svgfile>option, the game will just be printed out to the console. - For the Human player, you can enter moves in Standard Algebraic Notation (e.g. e4, Bxf3, or O-O-O) or UCI Notation (e.g. e2e4, g4f3, e1c1).
- You can't play ANN vs. ANN because you can only restore one TensorFlow session from the data at a time.