AWS MLRun multiple generative AI models on GPU using Amazon SageMaker multi-model endpoints with TorchServe and save up to 75% in inference costs
Host generative AI models on Amazon SageMaker multi-model endpoints using TorchServe and save on inference costs
AWS MLFine-tune Llama 2 for text generation on Amazon SageMaker JumpStart
Fine-tuning Llama 2 models for text generation with Amazon SageMaker JumpStart.
AWS MLIntelligently search Adobe Experience Manager content using Amazon Kendra
Configuring Amazon Kendra AEM connector to index and search AEM assets and pages, including filtering search results by user access
DuolingoA new tool for learning to read Japanese on Duolingo
Learn to read Japanese writing systems with Duolingo's new feature for reading kanji.
StripeThe creator economy goes global
Exploring the growth and international reach of the creator economy through the use of Stripe Connect by major creator platforms.
Snorkel AIHow we matured our ML-on-Kubernetes capabilities and saved on cloud costs
Optimizing ML-on-Kubernetes capabilities to reduce cloud costs by 40%
Snorkel AIHow we matured our ML-on-Kubernetes capabilities and saved on cloud costs
Improving margins and reducing cloud costs by maturing ML-on-Kubernetes capabilities.
DatabricksIntroducing Databricks Bengaluru Development Center
Introducing the new Databricks Bengaluru Development Center and its R&D teams in India
AWS MLOptimize deployment cost of Amazon SageMaker JumpStart foundation models with Amazon SageMaker asynchronous endpoints
Optimize deployment cost of Amazon SageMaker JumpStart foundation models by utilizing Amazon SageMaker asynchronous endpoints and reducing cold start time.
AWS MLHow Carrier predicts HVAC faults using AWS Glue and Amazon SageMaker
Predicting HVAC faults across large fleets of equipment using AWS Glue and Amazon SageMaker using ML and parallel data processing.
AWS MLBuild a generative AI-based content moderation solution on Amazon SageMaker JumpStart
Introducing a method for content moderation on image data using multi-modal pre-training and a large language model, enabling users to interact with images to confirm inappropriate content and generate structured JSON output.
Modular AIModular: Mojo🔥: A journey to 68,000x speedup over Python - Part 3
Exploring the journey to achieve a remarkable 68,000x speedup over Python through the use of Mojo🔥 in this third part of the blog series.