Samir Araujo

Get hands-on with AWS Trainium and PyTorch's TorchTitan to train models using a fully native PyTorch workflow. This workshop walks you through distributed training on Trainium, covering cluster setup, parallelism strategies, and performance optimization, all through familiar PyTorch APIs.

In this interactive chalk talk, we'll dive deep into advanced post-training techniques for large language models using AWS Trainium. Together, we'll architect solutions for supervised fine-tuning, knowledge distillation, and reinforcement learning, addressing real-world challenges in adapting trillion-parameter models.

While large language models (LLMs) have broad knowledge and versatility, tailoring these powerful models to specific domains unlocks new levels of understanding and performance, allowing them to enhance user experience in domain-focused or industry-vertical applications.

Amazon EC2 provides the broadest and deepest portfolio of instances for machine learning applications. From GPU-based high-performance instances such as P4 and G5, to Trn1 and Inf1 instances purpose-built with AWS silicon for best price performance, there’s a right instance for each of your machine learning workloads.

As a data scientist, choosing the right compute instance for your workload on AWS can be challenging. On AWS, you can choose from CPUs, GPUs, AWS Trainium, and Intel Habana Gaudi to accelerate training.

Machine learning is ranked the number-one “game changer” for the retail market segment by chief experience officers (CXOs), yet it’s only number eight on top spending priorities. So which scenarios are real?