Category: Amazon Machine Learning

Organizations are increasingly using multiple large language models (LLMs) when building generative AI applications. Although an individual LLM can be highly capable, it might not optimally address a wide range of use cases or meet diverse performance requirements. The multi-LLM approach enables organizations to effectively choose the right model for each task, adapt to different […]

In this post, we explore how to use Amazon Bedrock for synthetic data generation, considering these challenges alongside the potential benefits to develop effective strategies for various applications across multiple industries, including AI and machine learning (ML).

This post demonstrates how Amazon Bedrock Knowledge Bases can help you scale your data management effectively while maintaining proper access controls on different management levels.

Prompt caching, now generally available on Amazon Bedrock with Anthropic’s Claude 3.5 Haiku and Claude 3.7 Sonnet, along with Nova Micro, Nova Lite, and Nova Pro models, lowers response latency by up to 85% and reduces costs up to 90% by caching frequently used prompts across multiple API calls. This post provides a detailed overview of the prompt caching feature on Amazon Bedrock and offers guidance on how to effectively use this feature to achieve improved latency and cost savings.

In this blog, we discuss how to optimize prompting in Amazon Nova for the best price-performance.

Today, we’re excited to announce the general availability of these evaluation features in Amazon Bedrock Evaluations, along with significant enhancements that make them fully environment-agnostic. In this post, we explore these new features in detail, showing you how to evaluate both RAG systems and models with practical examples. We demonstrate how to use the comparison capabilities to benchmark different implementations and make data-driven decisions about your AI deployments.

In this post, we introduce a state-of-the-art method to fine-tune LLMs by reinforcement learning, reviewed the pros and cons of RLHF vs. RLAIF vs. DPO, and saw how to scale LLM fine-tuning efforts with RLAIF. We also see how to implement an end-to-end RLAIF pipeline on SageMaker using the Hugging Face Transformer and TRL libraries, and using either off-the-shelf toxicity reward models to align responses during PPO or by directly prompting an LLM to generate quantitative reward feedback during PPO.

Lumi is a leading Australian fintech lender empowering small businesses with fast, flexible, and transparent funding solutions. They use real-time data and machine learning (ML) to offer customized loans that fuel sustainable growth and solve the challenges of accessing capital. This post explores how Lumi uses Amazon SageMaker AI to meet this goal, enhance their transaction processing and classification capabilities, and ultimately grow their business by providing faster processing of loan applications, more accurate credit decisions, and improved customer experience.

OMRON Corporation is a leading technology provider in industrial automation, healthcare, and electronic components. In their Shaping the Future 2030 (SF2030) strategic plan, OMRON aims to address diverse social issues, drive sustainable business growth, transform business models and capabilities, and accelerate digital transformation. At the heart of this transformation is the OMRON Data & Analytics Platform (ODAP), an innovative initiative designed to revolutionize how the company harnesses its data assets. This post explores how OMRON Europe is using Amazon Web Services (AWS) to build its advanced ODAP and its progress toward harnessing the power of generative AI.

Ray is an open source framework that makes it straightforward to create, deploy, and optimize distributed Python jobs. In this post, we demonstrate the steps involved in running Ray jobs on SageMaker HyperPod.