This is a Fork from python-engineer/snake-ai-pytorch.

After studing the basics to get to this point, i want to play around with an increased paramater count for the game state, and a more in depth rewards table. I want to see if i cant create a more complex model that will yeild better results in longer games and avoid the common error of looping into itself unavoidably. More to come. WIP