An Extendible (General) Continual Learning Framework based on Pytorch - official codebase of Dark Experience for General Continual Learning

This repository contains model-free deep reinforcement learning algorithms implemented in Pytorch

(TNNLS) Prioritized Experience-Based Reinforcement Learning with Human Guidance for Autonomous Driving

Implementation of "Episodic Memory in Lifelong Language Learning"(NeurIPS 2019) in Pytorch

1'st Place approach for CVPR 2020 Continual Learning Challenge

Implementation of HindSight Experience Replay paper with Pytorch

Repository containing code for the paper "Meta-Learning with Sparse Experience Replay for Lifelong Language Learning".

[ 👾 ] ➡️ 💾 ➡️ { 🎮🕹️ } Extra Stable-Baselines3 buffer classes. Reducing RL memory usage drastically with minimal overhead.

Framework for developing Actor-Critic deep RL algorithms (A3C, A2C, PPO, GAE, etc..) in different environments (OpenAI's Gym, Rogue, Sentiment Analysis, Car Controller, etc..) with continuous and discrete action spaces.

RBDoom is a Rainbow-DQN based agent for playing the first-person shooter game Doom

Off-Policy Correction for Actor-Critic Algorithms in Deep Reinforcement Learning

Train an agent using RL to navigate (and collect bananas) in a large, square world

A multi agent reinforcement learning environment where two agents controlled by DRQNs play a custom version of the pursuit-evasion game.

Distributed RL platform with modified IMPALA architecture. Implements CLEAR, LASER V-trace modifications along with Attentive and Elite sampling experience replay methods.

M.Sc. thesis on Continual Learning for Non-Autoregressive Neural Machine Translation

RL based agent for atari games

Combining Experience Replay with Exploration by Random Network Distillation

A repository of Q-learning based Deep Reinforcement learning algorithms, including Linear DQN, DQN with experience reply, Dueling DQN and Double Dueling DQN. Mostly tested on Gym environments.

QWOP agent

Another Addition to the Pile of Deep Q Learning, Double DQN, PER, Dueling DQN Implementations