How to use reinforcement learning environment for data generation

How do I use Open AI gym environment and reinforcement learning toolbox to generate data for training a neural network?

Thank you.