Amazon SageMaker AI
AWS adds SOCI lazy loading to cut SageMaker AI job start times from minutes to seconds
AWS has introduced Seekable OCI (SOCI) indexing for Amazon SageMaker AI Training and processing jobs, a feature that lazy-loads container images so jobs can start in…
Exowatt fine-tunes small Qwen3 models on SageMaker to beat frontier models
Solar and thermal storage company Exowatt says it has built a repeatable pipeline on Amazon SageMaker AI that produces small, task-specific models with higher accuracy…
AWS shows multi-turn reinforcement learning on SageMaker AI with a Wordle agent
AWS has published a hands-on guide to Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI, using Wordle to show how to train a model on multi-step agent…
SageMaker lets teams write their own training loops without managing GPUs
Amazon SageMaker AI has added training session APIs to its serverless model customization, letting developers write their own fine-tuning loops in Python while AWS…
Deepgram's Nova-3 on AWS Trainium2 processes 11, 70% more audio per dollar
Deepgram and the AWS Forward Deployed Engineering team have ported the Nova-3 speech-to-text model to AWS Trainium2, and report that the setup processed 11, 70 percent…
Meta's Muse Spark Open lands in Amazon SageMaker JumpStart
Meta's Muse Spark Open, an open-weight model optimized for coding, is now available in Amazon SageMaker JumpStart. Users can deploy it to an endpoint in their own AWS…
AWS Details Two MoE Kernel Tweaks That Lift SageMaker AI Throughput on H200
AWS has described two kernel-level optimizations for fused Mixture-of-Experts (MoE) models on Amazon SageMaker AI, reporting a 2, 4 percent end-to-end output throughput…
AWS adds SageMaker inference skill for Kiro, Claude Code and Codex
Amazon SageMaker AI optimized generative AI inference now offers the aws-ai-ml skill, which gives coding agents such as Kiro, Claude Code and Codex expertise in…
AWS shows how to train a search agent with multi-turn RL on SageMaker AI
AWS has published a walkthrough of how to fine-tune a search agent with Amazon SageMaker AI multi-turn reinforcement learning (MTRL). The post describes training a…
Amazon Payments Uses Contextual Bandits to Personalize Its Acquisition Funnel
Amazon Payments has applied a multi-objective contextual multi-armed bandit on Amazon SageMaker AI to personalize a product acquisition funnel. According to AWS, a…
uniopen Tunes Amazon Nova 2 Lite to Meet Retail Moderation Targets
uniopen, the digital communication and membership platform from Taiwan's Uni-President Enterprises Group, adapted Amazon Nova 2 Lite to its own retail moderation…