Tencent / Phoenixgo

PhoenixGo is a Go AI program which implements the AlphaGo Zero paper "Mastering the game of Go without human knowledge". It is also known as "BensonDarr" and "金毛测试" in FoxGo, "cronus" in CGOS, and the champion of World AI Go Tournament 2018 held in Fuzhou China.

If you use PhoenixGo in your project, please consider mentioning in your README.

If you use PhoenixGo in your research, please consider citing the library as follows:

@misc{PhoenixGo2018,
  author = {Qinsong Zeng and Jianchang Zhang and Zhanpeng Zeng and Yongsheng Li and Ming Chen and Sifan Liu}
  title = {PhoenixGo},
  year = {2018},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/Tencent/PhoenixGo}}
}

Building and Running

On Linux

Requirements

• GCC with C++11 support
• Bazel (0.19.2 is known-good)
• (Optional) CUDA and cuDNN for GPU support
• (Optional) TensorRT (for accelerating computation on GPU, 3.0.4 is known-good)

The following environments have also been tested by independent contributors : here. Other versions may work, but they have not been tested (especially for bazel).

Download and Install Bazel

Before starting, you need to download and install bazel, see here.

For PhoenixGo, bazel (0.19.2 is known-good), read Requirements for details

If you have issues on how to install or start bazel, you may want to try this all-in-one command line for easier building instead, see FAQ question

Building PhoenixGo with Bazel

Clone the repository and configure the building:

$ git clone https://github.com/Tencent/PhoenixGo.git
$ cd PhoenixGo
$ ./configure

./configure will start the bazel configure : ask where CUDA and TensorRT have been installed, specify them if need.

Then build with bazel:

$ bazel build //mcts:mcts_main

Dependices such as Tensorflow will be downloaded automatically. The building process may take a long time.

Recommendation : the bazel building uses a lot of RAM, if your building environment is lack of RAM, you may need to restart your computer and exit other running programs to free as much RAM as possible.

Running PhoenixGo

Download and extract the trained network:

$ wget https://github.com/Tencent/PhoenixGo/releases/download/trained-network-20b-v1/trained-network-20b-v1.tar.gz
$ tar xvzf trained-network-20b-v1.tar.gz

The PhoenixGo engine supports GTP (Go Text Protocol), which means it can be used with a GUI with GTP capability, such as Sabaki. It can also run on command-line GTP server tools like gtp2ogs.

But PhoenixGo does not support all GTP commands, see FAQ question.

There are 2 ways to run PhoenixGo engine

1) start.sh : easy use

Run the engine : scripts/start.sh

start.sh will automatically detect the number of GPUs, run mcts_main with proper config file, and write log files in directory log.

You could also use a customized config file (.conf) by running scripts/start.sh {config_path}. If you want to do that, see also #configure-guide.

2) mcts_main : fully control

If you want to fully control all the options of mcts_main (such as changing log destination, or if start.sh is not compatible for your specific use), you can run directly bazel-bin/mcts/mcts_main instead.

For a typical usage, these command line options should be added:

• --gtp to enable GTP mode
• --config_path=replace/with/path/to/your/config/file to specify the path to your config file
• it is also needed to edit your config file (.conf) and manually add the full path to ckpt, see FAQ question. You can also change options in config file, see #configure-guide.
• for other command line options , see also #command-line-options for details, or run ./mcts_main --help . A copy of the --help is provided for your convenience here

For example:

$ bazel-bin/mcts/mcts_main --gtp --config_path=etc/mcts_1gpu.conf --logtostderr --v=0

(Optional) : Distribute mode

PhoenixGo support running with distributed workers, if there are GPUs on different machine.

Build the distribute worker:

$ bazel build //dist:dist_zero_model_server

Run dist_zero_model_server on distributed worker, one for each GPU.

$ CUDA_VISIBLE_DEVICES={gpu} bazel-bin/dist/dist_zero_model_server --server_address="0.0.0.0:{port}" --logtostderr

Fill ip:port of workers in the config file (etc/mcts_dist.conf is an example config for 32 workers), and run the distributed master:

$ scripts/start.sh etc/mcts_dist.conf

On macOS

Note: Tensorflow stop providing GPU support on macOS since 1.2.0, so you are only able to run on CPU.

Use Pre-built Binary

Download and extract CPU-only version (macOS)

Follow the document included in the archive : using_phoenixgo_on_mac.pdf

Building from Source

Same as Linux.

On Windows

Recommendation: See FAQ question, to avoid syntax errors in config file and command line options on Windows.

Use Pre-built Binary

GPU version :

The GPU version is much faster, but works only with compatible nvidia GPU. It supports this environment :

• CUDA 9.0 only
• cudnn 7.1.x (x is any number) or lower for CUDA 9.0
• no AVX, AVX2, AVX512 instructions supported in this release (so it is currently much slower than the linux version)
• there is no TensorRT support on Windows

Download and extract GPU version (Windows)

Then follow the document included in the archive : how to install phoenixgo.pdf

note : to support special features like CUDA 10.0 or AVX512 for example, you can build your own build for windows, see #79

CPU-only version :

If your GPU is not compatible, or if you don't want to use a GPU, you can download this CPU-only version (Windows),

Follow the document included in the archive : how to install phoenixgo.pdf

Configure Guide

Here are some important options in the config file:

• num_eval_threads: should equal to the number of GPUs
• num_search_threads: should a bit larger than num_eval_threads * eval_batch_size
• timeout_ms_per_step: how many time will used for each move
• max_simulations_per_step: how many simulations(also called playouts) will do for each move
• gpu_list: use which GPUs, separated by comma
• model_config -> train_dir: directory where trained network stored
• model_config -> checkpoint_path: use which checkpoint, get from train_dir/checkpoint if not set
• model_config -> enable_tensorrt: use TensorRT or not
• model_config -> tensorrt_model_path: use which TensorRT model, if enable_tensorrt
• max_search_tree_size: the maximum number of tree nodes, change it depends on memory size
• max_children_per_node: the maximum children of each node, change it depends on memory size
• enable_background_search: pondering in opponent's time
• early_stop: genmove may return before timeout_ms_per_step, if the result would not change any more
• unstable_overtime: think timeout_ms_per_step * time_factor more if the result still unstable
• behind_overtime: think timeout_ms_per_step * time_factor more if winrate less than act_threshold

Options for distribute mode:

• enable_dist: enable distribute mode
• dist_svr_addrs: ip:port of distributed workers, multiple lines, one ip:port in each line
• dist_config -> timeout_ms: RPC timeout

Options for async distribute mode:

Async mode is used when there are huge number of distributed workers (more than 200), which need too many eval threads and search threads in sync mode. etc/mcts_async_dist.conf is an example config for 256 workers.

• enable_async: enable async mode
• enable_dist: enable distribute mode
• dist_svr_addrs: multiple lines, comma sperated lists of ip:port for each line
• num_eval_threads: should equal to number of dist_svr_addrs lines
• eval_task_queue_size: tunning depend on number of distribute workers
• num_search_threads: tunning depend on number of distribute workers

Read mcts/mcts_config.proto for more config options.

Command Line Options

mcts_main accept options from command line:

• --config_path: path of config file
• --gtp: run as a GTP engine, if disable, gen next move only
• --init_moves: initial moves on the go board, for example usage, see FAQ question
• --gpu_list: override gpu_list in config file
• --listen_port: work with --gtp, run gtp engine on port in TCP protocol
• --allow_ip: work with --listen_port, list of client ip allowed to connect
• --fork_per_request: work with --listen_port, fork for each request or not

Glog options are also supported:

• --logtostderr: log message to stderr
• --log_dir: log to files in this directory
• --minloglevel: log level, 0 - INFO, 1 - WARNING, 2 - ERROR
• --v: verbose log, --v=1 for turning on some debug log, --v=0 to turning off

mcts_main --help for more command line options. A copy of the --help is provided for your convenience here

Analysis

For analysis purpose, an easy way to display the PV (variations for main move path) is --logtostderr --v=1 which will display the main move path winrate and continuation of moves analyzed, see FAQ question for details

It is also possible to analyse .sgf files using analysis tools such as :

• GoReviewPartner : an automated tool to analyse and/or review one or many .sgf files (saved as .rsgf file). It supports PhoenixGo and other bots. See FAQ question for details

FAQ

You will find a lot of useful and important information, also most common problems and errors and how to fix them

Please take time to read the FAQ