Vision-Based Continuous CartPole with DQN
Implementation of the continuous CartPole from OpenAI's Gym using only visual input for Reinforcement Learning control with Deep Q-Networks
You can change the discrete action space:
n_actions = 11
action_space = [
(-1.0), (-0.8), (-0.6), # Action Space Structure
(-0.4), (-0.2), (0), # (Steering Wheel, Gas, Break)
(0.2), (0.4), (0.6), # Range -1~1 0~1 0~1
(0.8), (1.0)
]