Using variables from in


I would like to make the training of the agent from dependent of variables from (for instance num_learning_iterations). What would be the best way to utilize the variables from in other task scripts such as Is there an example script that I can rely on?

Thank you in advance


You can access variables from via the VecTask class in For example, you can modify to set a variable in your task with self.vec_env.task.num_learning_iterations = num_learning_iterations. Or alternatively, you can also give your task access to the PPO object directly in python/rlgpu/’s train() by setting task.ppo = ppo, or task.num_learning_iterations = ppo_iterations.