Changes to make it work

[NikeHop]

I made a few changes, not all of them are relevant.

Overview of the changes:

main.py:

• Added tensorboard logging
• outsourced some of the code to utils.py to have a cleaner training file
• Added some CL arguments to be able to change the exact parameters of the environment

cross_circle_base.py:

• I contacted the authors and they gave me access to the environment they used. The implementation of your environment is more general so I just changed the default settings to the one they used, a link to their environment is at the beginning of the file

The other changes are as I mentioned in the GitHub issue before