AWS sagemaker documentation change

Service: sagemaker · 2026-06-28 · Documentation low

File: sagemaker/latest/dg/augmented-manifest.md

Summary

Added SageMaker Python SDK v3 code example for training models with augmented manifest files

Security assessment

The change only adds new SDK usage examples without any security context, vulnerability fixes, or security feature documentation. It demonstrates standard training configuration with ModelTrainer API.

Diff

diff --git a/sagemaker/latest/dg/augmented-manifest.md b/sagemaker/latest/dg/augmented-manifest.md
index 1e8c2b5be..e69fa9d65 100644
--- a//sagemaker/latest/dg/augmented-manifest.md
+++ b//sagemaker/latest/dg/augmented-manifest.md
@@ -135,0 +136,48 @@ The following shows how to train a model with an augmented manifest file using t
+SageMaker Python SDK v3
+    
+    
+    
+    import sagemaker
+    from sagemaker.train import ModelTrainer
+    from sagemaker.core.shapes import Channel, DataSource, S3DataSource
+    from sagemaker.train.configs import Compute, StoppingCondition, OutputDataConfig
+    
+    # Create a model object set to using "Pipe" mode.
+    compute = Compute(
+        instance_type='ml.p3.2xlarge',
+        instance_count=1,
+        volume_size_in_gb=50
+    )
+    
+    model_trainer = ModelTrainer(
+        training_image=training_image,
+        role=role,
+        compute=compute,
+        stopping_condition=StoppingCondition(max_runtime_in_seconds=360000),
+        training_input_mode='Pipe',
+        output_data_config=OutputDataConfig(s3_output_path=s3_output_location),
+        sagemaker_session=session
+    )
+    
+    # Create a train data channel with S3_data_type as 'AugmentedManifestFile' and attribute names.
+    train_data = Channel(
+        channel_name='train',
+        data_source=DataSource(
+            s3_data_source=S3DataSource(
+                s3_data_type='AugmentedManifestFile',
+                s3_uri=your_augmented_manifest_file,
+                s3_data_distribution_type='FullyReplicated',
+                attribute_names=['source-ref', 'annotations'],
+            )
+        ),
+        content_type='application/x-recordio',
+        input_mode='Pipe',
+        record_wrapper_type='RecordIO'
+    )
+    
+    # Train a model.
+    model_trainer.train(input_data_config=[train_data])
+
+SageMaker Python SDK v2 (Legacy)
+    
+