AWS nova documentation change
Summary
Updated product name from 'SageMaker AI Hyperpod' to 'SageMaker HyperPod' throughout the document. Added new section for MLflow monitoring.
Security assessment
Changes are branding updates (name standardization) and operational improvements (adding MLflow monitoring). No security vulnerabilities, encryption methods, or access controls were modified. The existing security context about S3 encryption with KMS keys remains unchanged.
Diff
diff --git a/nova/latest/nova2-userguide/nova-hp.md b/nova/latest/nova2-userguide/nova-hp.md index 02fa9244d..2ff1c0294 100644 --- a//nova/latest/nova2-userguide/nova-hp.md +++ b//nova/latest/nova2-userguide/nova-hp.md @@ -7 +7 @@ Compute requirements for Amazon Nova 2 modelsBest Practices for Amazon Nova cust -# Amazon Nova customization on SageMaker AI Hyperpod +# Amazon Nova customization on SageMaker HyperPod @@ -9 +9 @@ Compute requirements for Amazon Nova 2 modelsBest Practices for Amazon Nova cust -You can customize Amazon Nova models, including the enhanced Amazon Nova 2.0 models, using [ Amazon Nova recipes](./nova-model-recipes.html) and train them on Hyperpod. A recipe is a YAML configuration file that provides details to SageMaker AI on how to run your model customization job. SageMaker AI Hyperpod supports two types of services: Forge and Non-forge. +You can customize Amazon Nova models, including the enhanced Amazon Nova 2.0 models, using [ Amazon Nova recipes](./nova-model-recipes.html) and train them on Hyperpod. A recipe is a YAML configuration file that provides details to SageMaker AI on how to run your model customization job. SageMaker HyperPod supports two types of services: Forge and Non-forge. @@ -13 +13 @@ Hyperpod offers high-performance computing with optimized GPU instances and Amaz -Amazon Nova customization on SageMaker AI Hyperpod stores model artifacts including model checkpoints in a service-managed Amazon S3 bucket. Artifacts in the service-managed bucket are encrypted with SageMaker AI-managed AWS KMS keys. Service-managed Amazon S3 buckets don't currently support data encryption using customer-managed KMS keys. You can use this checkpoint location for evaluation jobs or Amazon Bedrock inference. +Amazon Nova customization on SageMaker HyperPod stores model artifacts including model checkpoints in a service-managed Amazon S3 bucket. Artifacts in the service-managed bucket are encrypted with SageMaker AI-managed AWS KMS keys. Service-managed Amazon S3 buckets don't currently support data encryption using customer-managed KMS keys. You can use this checkpoint location for evaluation jobs or Amazon Bedrock inference. @@ -19 +19 @@ Standard pricing can apply for compute instances, Amazon S3 storage, and FSx for -The following tables summarize the computational requirements for SageMaker AI Hyperpod and SageMaker AI training jobs training for Amazon Nova 2 models. +The following tables summarize the computational requirements for SageMaker HyperPod and SageMaker AI training jobs training for Amazon Nova 2 models. @@ -25,3 +25,3 @@ SFT (Full Rank) | 4 | P5.48xlarge | 32 | Full model fine-tuning | Nova 2 Lite -RFT on SageMaker AI Training Jobs (LoRA) | 2 | P5.48xlarge | 16 | Custom Reward Functions in your AWS Environment | Nova 2 Lite -RFT on SageMaker AI Training Jobs (Full Rank) | 4 | P5.48xlarge | 32 | 32K context length | Nova 2 Lite -RFT on SageMaker AI Hyperpod | 8 | P5.48xlarge | 64 | Default 8192 context length | Nova 2 Lite +RFT on SageMaker Training Jobs (LoRA) | 2 | P5.48xlarge | 16 | Custom Reward Functions in your AWS Environment | Nova 2 Lite +RFT on SageMaker Training Jobs (Full Rank) | 4 | P5.48xlarge | 32 | 32K context length | Nova 2 Lite +RFT on SageMaker HyperPod | 8 | P5.48xlarge | 64 | Default 8192 context length | Nova 2 Lite @@ -131 +131 @@ Don't use SFT when the gap is knowledge rather than behavior. It doesn't teach t - * [Creating a SageMaker AI HyperPod EKS cluster with restricted instance group (RIG)](./nova-hp-cluster.html) + * [Creating a SageMaker HyperPod EKS cluster with restricted instance group (RIG)](./nova-hp-cluster.html) @@ -133 +133 @@ Don't use SFT when the gap is knowledge rather than behavior. It doesn't teach t - * [Amazon SageMaker AI HyperPod Essential Commands Guide](./nova-hp-essential-commands-guide.html) + * [Amazon SageMaker HyperPod Essential Commands Guide](./nova-hp-essential-commands-guide.html) @@ -135 +135 @@ Don't use SFT when the gap is knowledge rather than behavior. It doesn't teach t - * [Nova Forge access and setup for SageMaker AI Hyperpod ](./nova-forge-hp-access.html) + * [Nova Forge access and setup for ](./nova-forge-hp-access.html) @@ -140,0 +141,2 @@ Don't use SFT when the gap is knowledge rather than behavior. It doesn't teach t + * [Monitoring HyperPod jobs with MLflow](./nova-hp-mlflow.html) +