All Off-policy for Single agent are unavailable (td3, sac etc.)

[Sgmltd]

Thank you very much to you and your team for developing such an amazing repository! Its open sourcing will greatly fill the gap in open-source simulation environments for drones!

While using the code for single-agent training, I encountered issues with algorithms like TD3 and SAC, all reporting the same error:

policy = ALGOS[cfg.algo.name.lower()](
TypeError: SACPolicy.__init__() got multiple values for argument 'device'

As a result, there are currently no available off-policy algorithms for single-agent training in this code repository, which we deeply regret.

Please excuse my presumption, but could you or your team possibly address this amidst your busy schedules?

I apologize for intruding on your work and life. Your help in resolving this matter would be greatly appreciated! Once again, thank you for your outstanding work!