AWS User Group Wrocław Workshop - AWS Deep Racer

We would like to invite you to a full-day workshop on Amazon DeepRacer, co-organized by us and the AWS User Group in Wrocław, in cooperation with AWS Poland. This event will take place on Thursday, June 6, 2024, at 9:00 AM at our DataArt office - plac Grunwaldzki 23, 50-365 Wrocław. 6 floor.

To participate, you have to fill out the Google Form linked here.

AWS DeepRacer is a 1/18th scale autonomous race car designed to get developers hands-on with reinforcement learning (RL). During the workshop, using AWS RoboMaker, we will create a simulation environment in which our reinforcement learning agent (i.e., the autonomous car) can learn to drive autonomously. The training process is facilitated using the AWS DeepRacer service.

Workshop Details:

• The workshop will be conducted in English.
• Duration: Up to 6 hours (with the possibility of extension if necessary)
• Capacity: Limited to 50 participants. On-site, we will divide participants into groups.
• AWS Accounts: Provided.
• Equipment: All necessary equipment will be prepared by AWS Poland.
• Requirements: Participants need to bring their own laptops and chargers.
• The workshop is designed for everyone, both technical and non-technical attendees. No prior knowledge of AWS DeepRacer is required.

AWS DeepRacer is an engaging way to get started with machine learning (ML). This service provides a cloud environment that allows customers to train autonomous car models on a virtual track. These models can be tested during the DeepRacer event, where they are deployed to 1/18th scale cars racing on a track that matches the virtual one.

The core element of training a model is writing a reward function, which is a simple if-else Python function. No prior coding experience is required, although it is beneficial. Example reward functions will be provided, so participants do not have to start from scratch. Participants are encouraged to train multiple models simultaneously.

Once a model is trained, it is evaluated on a virtual track. Participants then select the best model(s) for the physical race. After testing on the physical track, participants can further improve their models by tweaking the reward function and retraining.

Event Agenda:

• Presentation: 45 minutes on Machine Learning and DeepRacer.
• Short demonstration of the DeepRacer console and model training.
• Training and Racing: Participants will train and evaluate models on-site (the first model will be ready in about 1.5 hours) and race throughout the day.

Don't miss this unique opportunity to learn about machine learning and participate in exciting races!

See you there!