rickstaa/stable-learning-control

Check the difference between the LAC TF1 and SAC replay buffers

Closed this issue · 1 comments

Describe the bug
The replay buffer object in the LAC implementation (see pool.py#L75-L84) works different than the one in the spinningup SAC implementation. We need to keep this in mind when adding the LAC Torch version to the mlc framework.

Present in v0.5.0