Location
Badges
Activity
Ratings Progression
Challenge Categories
Challenges Entered
Robustness and teamwork in a massively multiagent environment
Latest submissions
A new benchmark for Artificial Intelligence (AI) research in Reinforcement Learning
Latest submissions
See Allgraded | 9052 | ||
graded | 8888 | ||
graded | 8880 |
Participant | Rating |
---|
Participant | Rating |
---|
Unity Obstacle Tower Challenge
Docker Image Not Building/Running Correctly
Almost 5 years agoI created the docker image using the guide and build.sh.
Running the docker command:
docker run --rm --env OTC_EVALUATION_ENABLED=true --env OTC_DEMO_EVALUATION=true --network=host -it obstacle_tower_challenge:latest ./run.sh
produces the following error:
root Traceback (most recent call last): File "run.py", line 1, in <module> from obstacle_tower_env import ObstacleTowerEnv File "/home/aicrowd/obstacle_tower_env.py", line 2, in <module> from PIL import Image ModuleNotFoundError: No module named 'PIL'
And as a result, my submission times out and my agent never runs.
My requirements.txt file has the following
git+git://github.com/Unity-Technologies/obstacle-tower-env@v2.2 aicrowd-repo2docker
I tried adding pillow and specific versions of pillow, but Pillow seems to get installed as a prereq for obstacle-tower-env as it also requires ml-agents.
Does anyone have any suggestions?
[Admin] Good News for Dopamine Lovers!
Almost 5 years agoIn which script do you override that? I am having difficulties uploading my 4.6GB observation file, but as of now my agent doesnβt seem to perform as if it were trained unless it loads the observation data.
Edit:
Search project wasnβt working for me, I found it in the last method, βunbundleβ, of dqn_agent.py
Submission Dir Structure/Files
Almost 5 years agoYes. I run into that problem too, even using Tower to push, which solved my initial LFS problems. Iβm not sure how to push. This is the error I JUST got once again:
Pushing to git@gitlab.aicrowd.com:HappySlice/submission-tower-challenge.git
Connection to gitlab.aicrowd.com closed by remote host.
batch response: Authentication required: Authorization error: http://gitlab.aicrowd.com/HappySlice/submission-tower-challenge.git/info/lfs/objects/batch
Check that you have proper access to the repository
error: failed to push some refs to βgit@gitlab.aicrowd.com:HappySlice/submission-tower-challenge.gitβ
Submission Dir Structure/Files
Almost 5 years agoPutting everything in the root of your project repo is the way to go. I overthought it. The steps they list are confusing, just put run.sh in the most top part of your repository
Submission Dir Structure/Files
Almost 5 years agoThe problem Iβm currently running into is the actual folder structure of the submission to correctly execute.
I am currently receiving an error during submission:
2019-04-28T16:27:12.67245541Z bash: /home/aicrowd/run.sh: No such file or directory
So should we have the run.sh file in:
<.your project root folder.>/home/aicrowd/
Or is home regarded the same as root?:
<.your project root folder.>/aicrowd/
I copied run.sh in both places and the same error is being thrown, so Iβm not sure what to try next.
Submission Dir Structure/Files
Almost 5 years agoI had problems with Git LFS working and I just followed this guide and created an entirely new repo on gitlab and the large files uploaded.
Docker Not Running Environment
Almost 5 years agoI built the docker image in the obstacle-tower-challenge folder. When I run the docker code provided for ./run.sh, the agent correctly starts.
obstacle-tower-challenge davoo$ docker run --env OTC_EVALUATION_ENABLED=true --network=host -it obstacle_tower_challenge:latest ./run.sh
root
INFO:mlagents_envs:Start training by pressing the Play button in the Unity Editor.
When I run this command to start the environment
docker run --env OTC_EVALUATION_ENABLED=true --env OTC_DEMO_EVALUATION=true --network=host -it obstacle_tower_challenge:latest ./env.sh
It just spits out what the console would spit out if I were to run env.sh then the console goes back to accepting input and the agent times out waiting for the Unity Editor to start.
`+ ENV_PORT=
- ENV_FILENAME=
- β[β -z ββ β]β
- ENV_PORT=5005
- β[β -z ββ β]β
- ENV_FILENAME=./ObstacleTower/obstacletower.x86_64
- touch otc_out.json
- APP_PID=10
- xvfb-run --auto-servernum ββserver-args=-screen 0 640x480x24β ./ObstacleTower/obstacletower.x86_64 --port 5005 2
- TAIL_PID=11
- wait 10
- tail -f otc_out.json`
I have no experience with docker. Any tips? Thanks.
Using dopamine trained model
Almost 5 years agoTo clarify. Does one just move the saved models to the base_dir and enter the command as if they were going to train except set the training steps to 1?
Using dopamine trained model
Almost 5 years agoHow do you load the model? Iβm not finding any references on how to do that
Docker Image Not Building/Running Correctly
Almost 5 years agoMohanty, is there an updated guide somewhere, or a good resource on this?
I added the environment.yaml file and now I am getting the following:
Any assistance is appreciated.
Thanks.