search

amazon-archives / aws-robomaker-sample-application-deepracer

Use AWS RoboMaker and demonstrate running a simulation which trains a reinforcement learning (RL) model to drive a car around a track
MIT License
112 stars 59 forks source link

Issue with AWS DeepRacer simulator #25

Closed billchenxi closed 5 years ago

billchenxi commented 5 years ago

Hi there, hope you are doing well. I have a horrible week using AWS DeepRacer to train my model. What my issue is that the simulator starts always at the (0,0) coordinate and not on the track I supposed to train. See the images below

Problem

and

problem2

After about 30 minutes, the car was still seeing only the grasses around. Is there something wrong with the simulator???

If there's a bug, please stop this gimmick, because we have to pay money for using the AWS!!!

Thanks, Bill

adam-dunc commented 5 years ago

Hi Bill, we are very sorry to hear you are having a bad experience with the DeepRacer.

From the screenshots, It looks like you are running this from the DeepRacer console, is that correct? We want to confirm you did not run the sample application in RoboMaker and you did not take the code here from github and modify/run it independently to narrow down the troubleshooting. Can you please confirm the exact steps you took to get to this scenario? This is the first time this has been reported and we've been unable to reproduce it.

We will also likely need your AWS account information to try to investigate further into the logs if we are unable to reproduce. If you'd like to send us any information about your account please email: ros-contributions@amazon.com

-Adam

madan96 commented 5 years ago

Just in case if you were using the local training launch file for this, I was encountering the same issue. In my case, I had forgot to set the MARKOV_PRESET_FILE environment variable to deepracer.py. Hope it helps :slightly_smiling_face:

nburek commented 5 years ago

@billchenxi let us know if @madan96's suggestion solves the issue for you. In the mean time, I'm going to close this issue as there is not enough information for us to reproduce it at the moment. Please reopen if you are still experiencing this issue or have additional information on how you were running the simulation that we can use to try to reproduce.