Comet Panels Contest

Comet Panel Contest

Help us build the largest visualization gallery for machine learning! Join thousands of data scientists around the world and submit your panel. The first 200 users to submit a Panel win a free t-shirt (or make a donation to your favorite charity). Submit your original ML-related Panel below!

If you need help with any of these steps, post your question on the Comet Slack Channel.

Steps to entering the contest:

Step 1: Log into and go to one of your existing projects, or create a new experiment.

Note that you’ll need to use a Project other than “general” (Uncategorized Experiments) in order to make your Project public, and the Project will need at least one experiment.

Step 2: Create a new Comet Panel, or alter an existing one.

Make sure that:

  • your Panel offers some utility to Data Science or Machine Learning
  • you have contributed enough originality if basing your panel on others’ work
  • you give credit to others where due

Some resources:

Step 3: Make your Project public.

Select Public under Project Visibility under Manage.

Step 4: Make your Panel public.

Step 5: Make sure your Panel is easily usable by others.

Make sure you:

  • give your Panel a good Description of what it does, and anything that is required to make it function.
  • describe any options and what they do.
  • provide a thumbnail image for the Panel (click the thumbnail under Description in the Code Editor).

Step 6: Save your Panel, and add it to your public Project.

Step 7: Make sure to save your view! Make your view the default by clicking the circled check mark next to the name of the view.

Step 8: To test to make sure everything is public, try opening your project from an incognito browser window to make sure others can see your work.

Step 9: Create a Share link to your Public Panel on your Public View. Add this link on the form below.

Step 10: Celebrate! You have contributed to the world’s largest visualization gallery!

It’s easy to get started

And it's free. Two things everyone loves.