AWS nova documentation change
Summary
Minor terminology updates: replaced 'SageMaker AI HyperPod' with 'SageMaker HyperPod' and changed section title from 'Fine-tuning' to 'Direct preference optimization (DPO)'
Security assessment
Changes are purely terminological updates with no security implications. No security-related content was added or modified, and no vulnerabilities are referenced.
Diff
diff --git a/nova/latest/userguide/nova-sft-1.md b/nova/latest/userguide/nova-sft-1.md index 39faf3855..20446236b 100644 --- a//nova/latest/userguide/nova-sft-1.md +++ b//nova/latest/userguide/nova-sft-1.md @@ -74 +74 @@ In general, larger datasets require fewer epochs to converge, while smaller data -The following is a recipe for full-rank SFT that's intended for you to quickly start an SFT job on a SageMaker AI HyperPod cluster. This recipe also assumes that you have connected to your SageMaker AI HyperPod cluster using the correct AWS credentials. +The following is a recipe for full-rank SFT that's intended for you to quickly start an SFT job on a SageMaker HyperPod cluster. This recipe also assumes that you have connected to your SageMaker HyperPod cluster using the correct AWS credentials. @@ -120 +119,0 @@ The following is a recipe for full-rank SFT that's intended for you to quickly s - val_check_interval: 100 @@ -361 +360 @@ Supervised fine-tuning (SFT) -Fine-tuning +Direct preference optimization (DPO)