Closed tldoan closed 5 years ago
Hi, Thank you for your interest in the project :) I've been wanting to try this for a while now, but unfortunately I'm now busy with other projects, so I'm afraid that it'll have to wait... Cheers
Hey,
Yesterday , I your code CEM on SAC (I implemented CEM on top of a pytorch code of SAC I found ) and for Ant-v2 it plateau at 1000 while CEM -TD3 seems to provides good results. So problably while do the reverse (implementing sac from your code). Is there any parameters I should carefully tune when putting CEM on SAC???
thanks again !
Hi again, The decay of the covariance for the CEM algorithm should be tuned carefully, but the one that I used for my experiments should work fine. Tell me if you get something interesting !
Hi,
thanks for releasing your code ! Btw, have you by any chance tried implementing your CEM method on SAC algorithm?
thanks !