Skip to content

Latest commit

 

History

History
 
 

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Instructions

  1. Install python requirements: pip install -r minigui/requirements.txt

  2. Download a model from our public bucket.

  3. Make sure the command at the top of serve.py actually runs and prints GTP engine ready; if not, something is wrong with the rest of the minigo setup, like virtualenv or similar.

  4. Compile the typescript. (Requires typescript compiler). Running cd minigui; tsc should find and compile the relevant files.

  5. Set your current working directory to minigo root and start the flask server. Replace $MODEL_PATH with the path to your downloaded model. Replace $BOARD_SIZE with either 9 or 19 depending on the size of the model you downloaded. Replace $PORT with a port number, e.g. 5001.

python minigui/serve.py --model=$MODEL_PATH --board_size=$BOARD_SIZE --port=$PORT
  1. open localhost:5001 (or whatever value you used for $PORT).

  2. The buttons in the upper right that say 'Human' can be toggled to set which color Minigo will play.

  3. By default, Minigui will use the Python Minigo engine. You can use the C++ engine by first compiling it (see the README), then passing --engine=cc when starting minigui/serve.py. You will need to pass a frozen GraphDef proto as the --model command line argument instead of the saved parameter data that the Python backend requires.

    Here's an example, note that we specify the board size when compiling the C++ engine binary, not when running the Minigui server:

bazel build --define=board_size=19 cc:main
python minigui/serve.py --port=8888 --model=saved_models/000483-indus-upgrade.pb --engine=cc