How to hold on the game waitting the trainning of an agent

[Ja1r0]

I am trying to use this game environment to train an agent by learning algorithm.But each trainning step need much time,so that the game enviroment can't recieve an action without delay.Actually the game will stack at one frame due to the delay of the action command.How should I pause the game and train my agent,and start it again when the train step is over?I am struggling for this issue for a long time.

[Ja1r0]

I found that it might be something wrong with my own learning algorithm.

[remi-blaise]

The author of this repo is not active anymore, so don't hope receiving an answer soon ^^ Feel free to implement this yourself. Warning: this game is very slow et not suitable for production