Reinforcement learning with ros robot
I'm very sorry to ask this question again
My case is using ubuntu18 and ROS melody At the same time, because my own hardware configuration is rtx3070. Therefore, the corresponding CUDA version is 11.1. As a result, Pytorch and tensorflow suitable for CUDA are no longer suitable for python2 seven
So my idea is to create a python3 environment by Anaconda (so that you can install the appropriate Python and tensorflow). But how to carry out reinforcement learning for the gazebo robot created in the main environment(with python2.7) in the python3 environment?