Skip to content

Latest commit

 

History

History
73 lines (51 loc) · 2.44 KB

README.md

File metadata and controls

73 lines (51 loc) · 2.44 KB

RL Gobang

Kickstart

bazel build -c opt //mcts:capi
pip install -r requirements.txt

Training

Training uses multiple self-play processes and a single training process. Each self-play process runs MCTS search with neural network oracle to generate self-play games. Then these games will be transmitted to the training process with IPC to improve the neural network oracle. The training process also features a evaluator to check whether the new neural network is better than the previous check point. The evaluation criterion is that A is better than B only if A wins B with white stone (defensive position). Only when a definitely stronger network has arisen, new check point will be saved.

The master executable controls the training and self-play processes. It decides the number of self-play processes, which is decisive to the generating speed. master always finishes immediately since it only creates and terminates the worker processes. The worker processes (training/self-playing) exists in the form of daemons. So it may need additional efforts to deploy the program to Windows systems.

# start training
python src/master.py start

# stop training
python src/master.py kill

Playing with Checkpoints

gobang_env is a GUI program to visualize the level of certain checkpoint. It also contains functions to save the playing history to images.

python src/gobang_env.py

Available checkpoints:

Achievements

  • Beats tito (an AI who achieved 3 first and 2 second place in Gomocup) with black stone.

VS tito with black stone

  • Gets high rank at the platform "微信小程序-欢乐五子棋腾讯版" (in progress).

Performance Notes

The following optimizations are extremely useful in the training process.

  1. Removing Python runtime in MCTS search with a native library (C++/Rust) brings a speedup of roughly 30.

  2. Multi-process can be utilized to accelerate the training procedure by the number of self-play processes, which is bottleneck of the whole system.

  3. Virtual loss is a critical optimization trick to batch the neural network inference during MCTS self-play. It can easily speedup self-play by about 6 times.

Paper

AlphaZero