print the state norm information: state_avg, state_std We don't suggest to use running stat state. We directly do normalization on state using the historical avg and std eg. `state = (state + self.neg_state_avg) * self.div_state_std` in `PreprocessEnv.step_norm()`
(self, neg_avg=None, div_std=None)
source not stored for this graph (policy: none)