AWS nova documentation change
Summary
Updated terminology from 'AI training jobs' to 'Training Jobs', fixed hyperlink formatting for 'recipe' documentation, added new training methods (Nova distillation, DPO), removed Reinforcement Fine-Tuning section, and clarified encryption key management terminology (changed 'SageMaker AI-managed KMS keys' to 'SageMaker-managed KMS keys').
Security assessment
The encryption clarification maintains existing security controls (service-managed keys) without introducing new security features or addressing vulnerabilities. Terminology changes and training method updates are documentation improvements without explicit security context. No evidence of patched vulnerabilities or new security capabilities.
Diff
diff --git a/nova/latest/userguide/nova-model-training-job.md b/nova/latest/userguide/nova-model-training-job.md index 4b27ae341..648df1ebd 100644 --- a//nova/latest/userguide/nova-model-training-job.md +++ b//nova/latest/userguide/nova-model-training-job.md @@ -5 +5 @@ -# Amazon Nova customization on SageMaker AI training jobs +# Amazon Nova customization on SageMaker Training Jobs @@ -7 +7 @@ -SageMaker AI training jobs is an environment that enables you to train machine learning models at scale. It automatically provisions and scales compute resources, loads training data from sources like Amazon S3, executes your training code, and stores the resulting model artifacts. +SageMaker Training Jobs is an environment that enables you to train machine learning models at scale. It automatically provisions and scales compute resources, loads training data from sources like Amazon S3, executes your training code, and stores the resulting model artifacts. @@ -9 +9 @@ SageMaker AI training jobs is an environment that enables you to train machine l -The purpose of training is to customize the base Amazon Nova model using your proprietary data. The training process typically involves steps to prepare your data, choose a [recipe](https://docs.aws.amazon.com/sagemaker/latest/dg/nova-model-recipes.html), modify configuration parameters in YAML files, and submit a training job. The training process will output trained model checkpoint in a service-managed Amazon S3 bucket. You can use this checkpoint location for evaluation jobs. Nova customization on SageMaker AI training jobs stores model artifacts in a service-managed Amazon S3 bucket. Artifacts in the service-managed bucket are encrypted with SageMaker AI-managed KMS keys. Service-managed Amazon S3 buckets don't currently support data encryption using customer-managed KMS keys. +The purpose of training is to customize the base Amazon Nova model using your proprietary data. The training process typically involves steps to prepare your data, choose a [recipe](./nova-model-recipes.html), modify configuration parameters in YAML files, and submit a training job. The training process will output trained model checkpoint in a service-managed Amazon S3 bucket. You can use this checkpoint location for evaluation jobs. Nova customization on SageMaker AI training jobs stores model artifacts in a service-managed Amazon S3 bucket. Artifacts in the service-managed bucket are encrypted with SageMaker AI-managed KMS keys. Service-managed Amazon S3 buckets don't currently support data encryption using customer-managed KMS keys. @@ -17 +17,5 @@ For best practices, see [Best Practices](./nova-forge-sft.html#best-practices). - * [Fine-tune Nova 2.0](./nova-fine-tune-2.html) + * [Fine-tune Nova 1.0](./nova-fine-tune-1.html) + + * [Amazon Nova distillation](./nova-distillation.html) + + * [Direct Preference Optimization (DPO)](./nova-dpo-smtj.html) @@ -23,2 +26,0 @@ For best practices, see [Best Practices](./nova-forge-sft.html#best-practices). - * [Reinforcement Fine-Tuning (RFT) with Amazon Nova models](./nova-reinforcement-fine-tuning.html) -