Amazon SageMaker Workshop > SageMaker Components for Kubeflow Pipelines Workshop > Running the Pipeline > Batch Transform Output
Batch Transform Output
The pipeline executed a batch transformation job for you, getting inferences for each record in the validation dataset. In the pipeline results page, click on the sagemaker-batch-transformation stage and view the Input/Output section.
Locate the batch_transform_ouput field and copy the S3 path. List that path to see and download the batch job output.
aws s3 cp s3://BUCKET/mnist_kmeans_example/output/valid-data.csv.out .
more valid-data.csv.out
You'll see a set of records that looks like this:
{"closest_cluster": 4.0, "distance_to_cluster": 6.285102367401123}
You can inspect the results to see how well the model performed on the validation data set.