Baseline for variance reduction in Policy Gradient Algorithm

Modular implementation of Vanila Policy Gradient (VPG) algorithm with baseline using an RNN policy.

Dependencies

• Python 2.7 or 3.5
• TensorFlow 1.10
• gym
• numpy
• tqdm progress-bar

Features

• Using a value function based baseline for reducing the variance in the vanila policy gradient algorithms
• Using an RNN policy for giving the action probabilities for a reinforcement learning problem
• Using a sampler that reshape the trajectory to be feed into an RNN policy
• Using gradient clipping to solve the exploding gradient problem
• Using GRU to solve the vanishing gradient problem

Usage

To train a model for Cartpole-v0:

$ python test_graph_pg.py 

To view the tensorboard

$tensorboard --logdir .