Update my reinforcement learning algorithm with the concept of A3C(Actor-Critic agents)

I have a simple (non-tensorflow) Q Algorithm implemented which takes updates the values in my table that fit the right target. It does so with one agent only which means that it can only run once at a time to update the values. I want to change this algorithm to have 3 working Agents at a time asynchronously to update the same Q table. Essentially, I want to implement the concept of A3C([login to view URL]) but in a simple Q-learning Algorithm. Developer must have experience in reinforcement learning in python before.