https://github.com/seungeunrho/minimalRL

Implementations of basic RL algorithms with minimal lines of codes! (pytorch based)
https://github.com/seungeunrho/minimalRL

a2c a3c acer ddpg deep-learning deep-reinforcement-learning dqn machine-learning policy-gradients ppo pytorch reinforce reinforcement-learning sac simple

Last synced: 15 days ago
JSON representation

Implementations of basic RL algorithms with minimal lines of codes! (pytorch based)

• Host: GitHub
• URL: https://github.com/seungeunrho/minimalRL
• Owner: seungeunrho
• License: mit
• Created: 2019-04-23T10:57:44.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2023-04-22T09:06:41.000Z (about 1 year ago)
• Last Synced: 2024-02-24T02:32:07.684Z (4 months ago)
• Topics: a2c, a3c, acer, ddpg, deep-learning, deep-reinforcement-learning, dqn, machine-learning, policy-gradients, ppo, pytorch, reinforce, reinforcement-learning, sac, simple
• Language: Python
• Homepage:
• Size: 60.5 KB
• Stars: 2,693
• Watchers: 50
• Forks: 442
• Open Issues: 21
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• my-awesome-stars - seungeunrho/minimalRL - Implementations of basic RL algorithms with minimal lines of codes! (pytorch based) (Python)
• awesome-stars - seungeunrho/minimalRL - Implementations of basic RL algorithms with minimal lines of codes! (pytorch based) (Python)

README

        
# minimalRL-pytorch

Implementations of basic RL algorithms with minimal lines of codes! (PyTorch based)

* Each algorithm is complete within a single file.

* Length of each file is up to 100~150 lines of codes.

* Every algorithm can be trained within 30 seconds, even without GPU.

* Envs are fixed to "CartPole-v1". You can just focus on the implementations.

## Algorithms

1. [REINFORCE](https://github.com/seungeunrho/minimalRL/blob/master/REINFORCE.py) (67 lines)

2. [Vanilla Actor-Critic](https://github.com/seungeunrho/minimalRL/blob/master/actor_critic.py) (98 lines)

3. [DQN](https://github.com/seungeunrho/minimalRL/blob/master/dqn.py) (112 lines,  including replay memory and target network)

4. [PPO](https://github.com/seungeunrho/minimalRL/blob/master/ppo.py) (119 lines,  including GAE)

5. [DDPG](https://github.com/seungeunrho/minimalRL/blob/master/ddpg.py) (145 lines, including OU noise and soft target update)

6. [A3C](https://github.com/seungeunrho/minimalRL/blob/master/a3c.py) (129 lines)

7. [ACER](https://github.com/seungeunrho/minimalRL/blob/master/acer.py) (149 lines)

8. [A2C](https://github.com/seungeunrho/minimalRL/blob/master/a2c.py) (188 lines)

9. [SAC](https://github.com/seungeunrho/minimalRL/blob/master/sac.py) (171 lines) added!! 

10. [PPO-Continuous](https://github.com/seungeunrho/minimalRL/blob/master/ppo-continuous.py) (161 lines) added!!

11. [Vtrace](https://github.com/seungeunrho/minimalRL/blob/master/vtrace.py) (137 lines) added!!

12. Any suggestion ...? 

## Dependencies

1. PyTorch

2. OpenAI GYM ( > 0.26.2 IMPORTANT!! No longer support for the previous versions)

## Usage

```bash

# Works only with Python 3.

# e.g.

python3 REINFORCE.py

python3 actor_critic.py

python3 dqn.py

python3 ppo.py

python3 ddpg.py

python3 a3c.py

python3 a2c.py

python3 acer.py

python3 sac.py

```