James Jeun

Amazon EC2 provides the broadest and deepest portfolio of instances for machine learning (ML) applications. From P3 instances that provide the best performance for ML training to G4 instances for inference and Inf1 instances that provide the lowest-cost ML inference in the cloud, there’s a right instance for each of your ML workload needs.

In this session, the Facebook AI team discusses its major machine learning models and workloads and the infrastructure challenges it faced with large-scale distributed training. They share details of how they tested these ML workloads on AWS infrastructure and the results of this benchmarking.

The key barriers to the wider adoption of deep neural networks on industrial-sized datasets are the time and resources required to train them. Data scientists and machine learning engineers continue to demand shorter training times, as this provides them with increased agility to try new algorithms and optimization techniques.