simple_RL_framework_for simulation

Implementations of some basic RL algorithms with simple codes in Pytorch.

Features:

• Algorithms:
  • Proximal Policy Optimization (PPO)
  • PPO with LSTM integration (PPO-LSTM)
  • Soft Actor-Critic (SAC)
• Action Spaces:
  • Supports both continuous and discrete action spaces.
• Modularity:
  • Easily adaptable to other simulation environments by modifying the Set_env class in environment.py.

Run code

If you want to implement RL algorithms for other simulation environments, you just need to revise "Set_env" class in "environment.py" file.

How to run code :

python main.py

Algorithms Implemented

Command	Description
1. PPO	Proximal Policy Optimization Algorithms
2. PPO-LSTM (recurrent PPO)
3. SAC	Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

Dependencies

Pytorch Numpy