Open PRs:

1. Bipedal walker using Soft Actor-Critic
2. A small bug fix for SAC agent, while loading a pretrained model

Merged PRs:

1. DoubleDQN doesn’t utilize DoubleQLearning
2. Addition of q_networks
3. Removing terminal state for Pendulum-v0
4. Changes in parameter description for activation functions in ANN module
5. Changes in the Pendulum-v0 implementation
6. Forward function update for softplus_function.hpp
7. Addition of Normal Distribution to ANN module
8. Accessor and Mutator for action in q_learning
9. Added cartpole-dqn notebook
10. Addition of DuelingDQN and Noisy linear layer
11. Addition of Noisy DQN to QLearning
12. N-step learning for DQN
13. Adding support to use gym_tcp_api for training agents in q_learning
14. Doc changes for q_learning
15. Changes corresponding to mlpack repo in cartpole_dqn
16. Changes to the network initialization methods in Q_learning networks
17. Adding notebook for solving acrobot using dqn
18. Adding notebook for solving mountainCar using DQN
19. Added the pendulum notebook
20. Addition of Categorical DQN
21. Added LunarLander-v2 env with dqn notebook
22. Addition of Soft Actor-Critic to RL methods
23. Adding SAC example for pendulum env
24. Added support for multiple actions in action space for Soft Actor Critic