I am trying to make an RL model for reinforcement learning, and the environment is fairly simple and a 10×10 grid. During the training phase, it reaches the goal. But when I try to test this, The agent takes one direction and goes straight and hits a wall or obstacle. It won’t change direction. And this goes on in a loop. My reward function is based on distance from the current position to the goal and penalties are given for moving away from the goal and hitting anything. The following given is my model class.