Buidling neural network to solve OpenAI's cart pole balancing environment.

This environment corresponds to the version of the cart-pole problem described by Barto, Sutton, and Anderson in "Neuronlike Adaptive Elements That Can Solve Difficult Learning Control Problem". A pole is attached by an un-actuated joint to a cart, which moves along a frictionless track. The pendulum is placed upright on the cart and the goal is to balance the pole by applying forces in the left and right direction on the cart¹

Since the goal is to keep the pole upright for as long as possible, a reward of +1 for every step taken, including the termination step, is allotted. The threshold for rewards is 50.

All observations are assigned a uniformly random value in (0.0, 2.0)

• It has 6 fully connected layers followed by dropout layers with categorical cross entropy loss and Adam optimizer with a learning rate of 1e-3.

To run CartPole locally perform the following steps (requires Python 3.7+):

1. git clone https://github.com/Suraj-Bhor/CartPole.git
2. cd Cartpole
3. pip install -r requirements.txt
4. python cartpole.py

To visualize graphs the model saves the tensorboard checkpoints to /tmp/tflearn_logs/. Once the model is trained you can visualize the results buy issuing the following command.

• tensorboard --logdir='/tmp/tflearn_logs'

1. Sentdex
2. Gym Library
3. Tensorboard
4. Spinning up OpenAI