How to write aicrowd.json to associate with team?

Unity Obstacle Tower Challenge
#1

Now I have set up an repository, I submit my entry, and the pipeline passed. But I can’t find our name in the LeaderBoard. Even the last one, I will find it! But I can’t find anything about I or team name or team leader name.
my aicrowd.json is:
{
“challenge_id” : “unity-obstacle-tower-challenge-2019”,
“grader_id”: “unity-obstacle-tower-challenge-2019”,
“authors” : [“xxx”],
“description” : “Random Obstacle Tower agent”,
“gpu”: true
}
what I should fill in the authors? team name? team leader’s name? or all team member’s name?

#2

Hi @yangjob,

Did you do a tag push ?
I cant see your submission in the submissions list or on the leaderboard.

Instructions for submission are available here : https://github.com/Unity-Technologies/obstacle-tower-challenge/#submitting

Cheers,
Mohanty

#3

Thanks.I am going to try. Should I create a new tag when I commit new things every time?
And what I should fill in the authors? Team name? I have tried team name, my name and our team leader’s name.

#4

@yangjob: The evaluator picks up your username from the account which pushes the tag.
And you can commit as you please when working on the code. When you wish to submit, you should create and push a tag. Each new tag is a submission.

#5

I have did a tag push. But nothing happened. Is there any delay on the leaderboard since submssion?

#6

Does the pipeline passed mean submission succeed? I think the official rules is ambiguous. Perhaps my english is bad, I cant understand some about submission.

#7

When you do a tag push, a corresponding issue is created for your submission : https://gitlab.aicrowd.com/yangjob/obstacle-tower-challenge/issues/4

#8

Oh, thank you very much. I am just an assistant for my senior in my team in this competition. I’m responsible for submission, collect information and collect data set of image process(classifier). However, the submission bother me a lot. Thank you! If you are convenient, I would like to ask you later.

#9

I got it! In the evaluation, the error of image_build_failed appeared several times. Finally, I try to delete the environment.yml, then I succeed. Now I need to import checkpoints which store tensorflow args trained by Dopamine rainbow to run.py. How to use the checkpoints?