HEC_80629A_H2021

Learning Project,

How to use (DQN family)

in the command line

python run.py --conf configurations/lunarlander_dqn.json

in Colab, download the code

!rm -rf HEC_80629A_H2021/
!git clone https://github.com/qzhou-2020/HEC_80629A_H2021.git

link you google drive

from google.colab import drive
drive.mount('/content/drive')

install the package

!apt install python-opengl
!apt install ffmpeg
!apt install xvfb
!apt install imagemagick

!pip install box2d
!pip install gym[box2d]
!pip install pyvirtualdisplay

add code path to the sys.path

import sys
sys.path.insert(0, "HEC_80629A_H2021")

run the code

%run HEC_80629A_H2021/run.py --conf HEC_80629A_H2021/configurations/lunarlander_ddqn_per.json

make gif or mp4

%run HEC_80629A_H2021/eval_policy.py saved_policy_dir env_name --save --save-to filename.gif

How to use (DDPG and PPO)

Upload Jupyter Notebooks (ddpg.ipynb, PPO_pytorch_continous.ipynb under "policy_based models" folder) to Colab and execute.

Demo

Good one:

Bad one:

Interesting one: