Anymal-C navigation environment for Spot robot

[saa-97]

Hello everyone, I am trying to train a navigation policy for the spot robot. As a reference, I am using the anymal-c navigation environment. I already have a pre trained policy which basically tells the spot robot how to stand up and walk. I trained it using the velocity environment for the spot.

As seen in the navigation environment for anymal-c, a pretrained policy is used to train for navigation. However, no matter how much I try to implement this, the robot does not even utilize the pre trained policy. It's like the pretrained policy has no effect on the training. I have searched for a solution/better explanation for the navigation environment, but sadly there is none. Could anyone please explain to me what I might be doing wrong? Or how to correctly use the navigation environment for training other quadrupeds? Thank you!!