RL From Scratch

Reinforcement Learning From Scratch

Collection of RL algorithms implemented from scratch in PyTorch. From Q-Learning to Multi-Agent RL, explore fundamental and advanced reinforcement learning concepts.

Filter by category:
A2C (A2C)

A2C (A2C)

Implementation of A2C reinforcement learning algorithm
DDPG

DDPG

Implementation of DDPG reinforcement learning algorithm
DQN Frozenlake

DQN Frozenlake

Implementation of DQN-FrozenLake reinforcement learning algorithm
DQN Lunar

DQN Lunar

Implementation of DQN-Lunar reinforcement learning algorithm
DQN Taxi

DQN Taxi

Implementation of DQN-Taxi reinforcement learning algorithm
DQN Atari

DQN Atari

Implementation of DQN-atari reinforcement learning algorithm
DQN

DQN

Implementation of DQN reinforcement learning algorithm
Duel DQN

Duel DQN

Implementation of Duel-DQN reinforcement learning algorithm
Flappybird PPO

Flappybird PPO

Implementation of FlappyBird-PPO reinforcement learning algorithm
Frozen Lake

Frozen Lake

Implementation of Frozen-Lake reinforcement learning algorithm
Imitation Learning

Imitation Learning

Implementation of Imitation Learning reinforcement learning algorithm
MARL

MARL

Implementation of MARL reinforcement learning algorithm
IPPO

IPPO

Implementation of IPPO reinforcement learning algorithm
MAPPO

MAPPO

Implementation of MAPPO reinforcement learning algorithm
Self Play

Self Play

Implementation of Self Play reinforcement learning algorithm
PPO

PPO

Implementation of PPO reinforcement learning algorithm
Atari

Atari

Implementation of Atari reinforcement learning algorithm
MuJoCo

MuJoCo

PPO on MuJoCo benchmark
REINFORCE

REINFORCE

Implementation of REINFORCE reinforcement learning algorithm
RND

RND

Implementation of RND reinforcement learning algorithm
SAC

SAC

Implementation of SAC reinforcement learning algorithm
TD3

TD3

Implementation of TD3 reinforcement learning algorithm