SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.06135 · ICML · 2021

DouZero: Mastering DouDizhu with Self-Play Deep Reinforcement Learning

Ji Liu, Daochen Zha, Wenye Ma, Xia Hu, Jingru Xie, Sheng Zhang, Xiangru Lian

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 6 functions out of this paper's own repositories and ran 3 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
kwai/DouZero canonical 2 of 3
EdwardPooh/douzero-resnet-2.0 extension 1 of 1
Liberations/DouZero_For_HLDDZ_FullAuto extension 0 of 2
FunctionStatusWhere it lives
FarmerLstmModel Ran kwai/DouZero/douzero/dmc/models.py
code served (permissive licence) · get_code("3998e0632bc7b7de")
Model Ran kwai/DouZero/douzero/dmc/models.py
code served (permissive licence) · get_code("dfccfb88738de17d")
get_latest_model Ran EdwardPooh/douzero-resnet-2.0/Douzero_Resnet/auto_test.py
pointer only (licence: GPL-3.0) · get_code("aefba852d67c0aec")
LandlordLstmModel Not yet run kwai/DouZero/douzero/dmc/models.py
code served (permissive licence) · get_code("d2b55c75f6fdae82")
compute_loss Not yet run Liberations/DouZero_For_HLDDZ_FullAuto/douzero/dmc/dmc.py
code served (permissive licence) · get_code("7a21191505af1ea4")
learn Not yet run Liberations/DouZero_For_HLDDZ_FullAuto/douzero/dmc/dmc.py
code served (permissive licence) · get_code("ee16d8c3495cc1e3")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Games are abstractions of the real world, where artificial agents learn to compete and cooperate with other agents. While significant achievements have been made in various perfect-and imperfectinformation games, DouDizhu (a.k.a. Fighting the Landlord), a three-player card game, is still unsolved. DouDizhu is a very challenging domain with competition, collaboration, imperfect information, large state space, and particularly a massive set of possible actions where the legal actions vary significantly from turn to turn. Unfortunately, modern reinforcement learning algorithms mainly focus on simple and small action spaces, and not surprisingly, are shown not to make satisfactory progress in DouDizhu. In this work, we propose a conceptually simple yet effective DouDizhu AI system, namely DouZero, which enhances traditional Monte-Carlo methods with deep neural networks, action encoding, and parallel actors. Starting from scratch in a single server with four GPUs, DouZero outperformed all the existing DouDizhu AI programs in days of training and was ranked the first in the Botzone leaderboard among 344 AI agents. Through building DouZero, we show that classic Monte-Carlo methods can be made to deliver strong results in a hard domain with a complex action space. The code and an online demo are released 1 with the hope that this insight could motivate future work.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2106.06135")
get_code_for_paper("2106.06135")
have("2106.06135")

Connect an agent — have() is free.