Closed FelipeMartins96 closed 3 years ago
When implemented in the old environment, the intention was to clip rewards when something unexpected occurred and the value was too big, on the code as of now it is clipping normal rewards
https://github.com/robocin/envs/blob/5f422179ca24fdf9eca0bf53fe02f2af1a4c6f21/envs/rc_gym/vss/env_3v3/vss_gym_3v3.py#L194
maybe the same thing occurs here https://github.com/robocin/envs/blob/b4e98159dc03a604e7413bb280e46f7f69199cd9/envs/rc_gym/vss/env_3v3/vss_gym_3v3.py#L179
When implemented in the old environment, the intention was to clip rewards when something unexpected occurred and the value was too big, on the code as of now it is clipping normal rewards
https://github.com/robocin/envs/blob/5f422179ca24fdf9eca0bf53fe02f2af1a4c6f21/envs/rc_gym/vss/env_3v3/vss_gym_3v3.py#L194