Already trained model and rewards

Hello I have a quenstion about rewards and a trained model.

Is a trained model, with a brain attached to it, uses the rewards (AddReward() function) from training script or rewards are used only in training process?

I mean, if I attach the brain and remove all AddReward() calls, will the model still work?

Thanks,

It’ll work - rewards are only relevant for training. You can safely remove them at inference (although keeping them in won’t cause any issues either way).

Nice!
Thanks mbaske for the answer.