Recurrent Deterministic Policy Gradient (RDPG)

Overview

PyTorch implementation of Recurrent Deterministic Policy Gradient from the paper Memory-based control with recurrent neural networks

Requirements

$ python -m pip install --upgrade pip
$ pip install -r requirements.txt

Training:

$ python main.py --env Pendulum-v0 --max_episode_length 1000 --trajectory_length 10 --debug

Memory-based control with recurrent neural networks <https://arxiv.org/abs/1512.04455>
Continuous control with deep reinforcement learning <https://arxiv.org/abs/1509.02971>
DDPG implementation using PyTorch <https://github.com/ghliu/pytorch-ddpg>
PyTorch-RL <https://github.com/jingweiz/pytorch-rl>

Name		Name	Last commit message	Last commit date
Latest commit History 17 Commits
.gitignore		.gitignore
README.rst		README.rst
TODO.md		TODO.md
agent.py		agent.py
evaluator.py		evaluator.py
main.py		main.py
memory.py		memory.py
model.py		model.py
normalized_env.py		normalized_env.py
random_process.py		random_process.py
rdpg.py		rdpg.py
requirements.txt		requirements.txt
util.py		util.py