Model convergence steps

[AnujPatil110377]

How many steps should I expect the model to converge using the default SAC setup?
Thanks!!

[AnujPatil110377]

I’ve been training the agent using the SAC policy for 500,000 iterations, but it still consistently collides with obstacles. Additionally, I’ve noticed that the goal occasionally spawns either inside obstacles or outside the environment boundaries. Could you please help me troubleshoot these issues?