Join an AWS DeepRacer League Virtual Circuit race - AWS DeepRacer

Join an AWS DeepRacer League Virtual Circuit race

In this section, learn how to use the AWS DeepRacer console to submit your trained model to a Virtual Circuit race.

To enter the AWS DeepRacer League Virtual Circuit

  1. Sign in to the AWS DeepRacer console.

  2. From the main navigation pane, choose AWS Virtual Circuit.

  3. On the AWS Virtual Circuit page, under the Open races section, choose Enter race.

  4. If this is your first time participating in an AWS DeepRacer League racing event, set your alias in Racer name under AWS DeepRacer League racer name.

  5. Under Choose model, select the model you want to use from the Model list. Ensure that your model has been trained to handle the track shape.

  6. If this is your first time participating in an AWS DeepRacer League event, under League requirements, select your Country of residence. Once you select your country of residence and submit your first model, it is locked in for the racing season and will be verified when prizes are awarded. Then, accept the terms and conditions by selecting the checkbox.

  7. Choose Enter race to complete the submission. The submission quota for each race is 50.

    After your model is submitted, the AWS DeepRacer console starts its evaluation. The process can take up to 10 minutes.

  8. On the race page, review the race details.

  9. On the race page, note your submission status under your racer name.

  10. On the race page, view the ranking list on the leaderboard to see how your model compares with others.

    If your model doesn't finish the track in three consecutive trials, it is not included in the ranking list on the leaderboard. Your leaderboard ranking reflects your best performing submission. You also receive a national and regional season standings to gauge where you rank amongst other racers in your country and region.

    After you submit a model, try improving its performance by refining your reward function and iterating on your model. You can also train a new model with a different algorithm or action space. Learn, adjust, and race again to increase your chances for rewards.