A code reimplementation of DeepMind's "Multiagent Cooperation and Competition with Deep Reinforcement Learning" with Tensorflow