Fix to use float64 actions for off policy algorithms

[tobirohrer]

Description

Using float64 actions in continuous off-policy algorithms like TD3 and DDPG caused an error.
The core of the issue is how ReplayBuffer and RolloutBuffer are storing actions. The RolloutBuffer for On-Policy algorithms always casts the actions to float32. The ReplayBuffer on the other hand doesn´t. This MR fixes this issue for now by casting actions to float32 in the ReplayBuffer as well.

Motivation and Context

closes #1145

• I have raised an issue to propose this change (required for new features and bug fixes)

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation (update in the documentation)

Checklist

• I've read the CONTRIBUTION guide (required)
• I have updated the changelog accordingly (required).
• My change requires a change to the documentation.
• I have updated the tests accordingly (required for a bug fix or a new feature).
• I have updated the documentation accordingly.
• I have opened an associated PR on the SB3-Contrib repository (if necessary)
• I have opened an associated PR on the RL-Zoo3 repository (if necessary)
• I have reformatted the code using make format (required)
• I have checked the codestyle using make check-codestyle and make lint (required)
• I have ensured make pytest and make type both pass. (required)
  Disclaimer: The test test_vec_envs.py::test_render[SubprocVecEnv] fails on my local machine. But it fails on the master branch too.
• I have checked that the documentation builds using make doc (required)

[araffin]

LGTM, thanks =)
(and sorry for the delay)