Instructions for generating data, training supervised learning for agent before RL

[AhmedKhaled945]

Hello, was wondering if there were instructions that can facilitate supervised learning stage before RL, will be very useful in games like doudizho and mahjong,

Thanks in advance.

[daochenzha]

@AhmedKhaled945 Thanks for the suggestion. It needs some effort because we didn't consider SL at the beginning. I will be working for it later when I got time.

[AhmedKhaled945]

Perfect, thanks so much for this, wanted to ask how can i register a new model? i am nearly finishing a rule based model for sichuan mahjong.

[daochenzha]

@AhmedKhaled945 You may register the model here. See https://github.com/datamllab/rlcard/blob/master/rlcard/models/__init__.py

After registration, you can use

from rlcard import models
models.load($MODEL_NAME$)

to load the model

[wanyubin123456]

完美，非常感谢，我想问我如何注册一个新模型？我几乎要完成基于规则的四川麻将模型。

Hello, have you finished your Sichuan mahjong model? I'm from Tsinghua University, China. I'm following up this project. I'm looking forward to your reply!