search

datamllab / rlcard

Reinforcement Learning / AI Bots in Card (Poker) Games - Blackjack, Leduc, Texas, DouDizhu, Mahjong, UNO.
http://www.rlcard.org
MIT License
2.87k stars 619 forks source link

Q: Player, Agent, Model #305

Closed KingKnecht closed 9 months ago

KingKnecht commented 9 months ago

Hi guys,

I'm getting really confused about Player, Model and Agent. I'm trying to understand the DMC-Doudizhu example. The model.py contains an agent which implements "step" and "step_eval" like other player classes. When I check "games/doudizhu" there is also a player class with "step" and "step_eval" functions.

What is the purpose of the player class? When are those classes/functions used? Is there a difference between learning and then playing against a trained model? Are there different classes used? Is there some documentation how this stuff is related to each other? A picture would be nice ;)

Thx

emanueleiacca commented 5 months ago

Hi can I send you a p.m. ? I have problems trying to understand how to play with the trained model after being trained. Thank you very much