Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

DQNcar in PythonClient is not working (will not converge) #1009

Closed
QiyuZ opened this issue Apr 25, 2018 · 2 comments
Closed

DQNcar in PythonClient is not working (will not converge) #1009

QiyuZ opened this issue Apr 25, 2018 · 2 comments

Comments

@QiyuZ
Copy link

QiyuZ commented Apr 25, 2018

There is an example in PythonClient which applies DQN to run the car. However, I ran it for a long time and found it will not converge. Does anyone success?

@mitchellspryn
Copy link
Contributor

mitchellspryn commented Apr 25, 2018

Not sure about the tutorial on this repo, but you may try ours here.

@sytelus sytelus added the ML-RL label Oct 17, 2018
@madratman madratman added the bug label Aug 13, 2020
@zimmy87 zimmy87 changed the title DQNcar in PythonClient is not work (will not converge) DQNcar in PythonClient is not working (will not converge) Jan 7, 2021
@zimmy87
Copy link
Contributor

zimmy87 commented Jan 15, 2021

Hi @QiyuZ, thank you for reporting this issue. This script was recently removed in #3215 and replaced with a new script at \AirSim\PythonClient\reinforcement_learning\dqn_car.py that uses the new OpenAI gym wrapper. I tested the new script locally and it runs for me. Can you retest with the new script and report back if you see any issues?

@zimmy87 zimmy87 closed this as completed Jan 15, 2021
# for free to join this conversation on GitHub. Already have an account? # to comment
Projects
None yet
Development

No branches or pull requests

6 participants