Hi, I have a question regarding the reward function design.
How do you choose the weights for each component of the reward?
Are there any guidelines, heuristics, or tuning strategies you follow ?
Understanding your approach would help in customizing or extending the reward for related tasks.
Thanks in advance!