ericgjackson / slumbot2019

Implementations of CFR for solving a variety of Holdem-like poker games
MIT License
133 stars 31 forks source link

How to train a slumbot for heads-up no-limit Texas Holdem? #2

Closed kli-casia closed 5 years ago

kli-casia commented 5 years ago

Dear @ericgjackson I am wondering how to use your code to train a bot to play heads-up no-limit Texas Holdem (like this one www.slumbot.com) There are lot of code in this repo, I want to have an intuitive understanding of the project by training a heads-up no-limit Texas Holdem bot step by step. Thank you very much! O(∩_∩)O

ericgjackson commented 5 years ago

As you say, there is a lot of code in the repository, which can make it difficult to get started. The code base is trying to be very full featured and general, which tends to obscure some of the basics. A couple of ideas:

1) Start with a simpler code base. You can find Kevin Waugh's sample code for solving a simple poker game here: http://poker.cs.ualberta.ca/open_cfr.html. You can find Rich Gibson's code here: https://github.com/rggibson/open-pure-cfr.

2) If sticking with this repository, maybe trace through what happens with a simple game. The README.md gives steps for solving a game I call "ms1f3". (It's a simple two street Holdem variant with three cards on the flop.) Focus on how regrets are updated and the "sumprobs". The relevant code will mostly be in vcfr.cpp and cfr_utils.cpp.

You might want to read about CFR first, although I don't know of a good introduction to point you to. Some of the published papers might be too technical and too theoretical to start with.

On Mon, Jun 17, 2019 at 5:01 AM Kai Li notifications@github.com wrote:

Dear @ericgjackson https://github.com/ericgjackson I am wondering how to use your code to train a bot to play heads-up no-limit Texas Holdem (like this one www.slumbot.com)? http://www.slumbot.com%EF%BC%89%EF%BC%9F There are lot of code in this repo, I want to have an intuitive understanding of the project by training a heads-up no-limit Texas Holdem bot step by step. Thank you very much! O(∩_∩)O

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ericgjackson/slumbot2019/issues/2?email_source=notifications&email_token=ABUMRNOVIQKSOW5OC3EL4U3P254IDA5CNFSM4HYV5Z22YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GZ3V27Q, or mute the thread https://github.com/notifications/unsubscribe-auth/ABUMRNMSZ7RWEVMVKEYPUV3P254IDANCNFSM4HYV5Z2Q .

kli-casia commented 5 years ago

Thanks,I will check ms1f3 first