How to do the training and load results?
Run the test_env script, will see a console message of the current state of the training
Run the load_training copy file and it will open the pusher_ppo1 file where it contains all the weights
Open the custom_arm_env file. This does not run and is opened by other files. The rewards and the setup of steps and observations lie here
This is the “urdf” of mujoco and will be edited for the actual robot or the pickup objects inside the world
type mjpython mujoc.py in the console while the environment gym is activated in conda in order to play with the joint values manually
