0

Recently, while using a known-good endpoint configuration, I keep on receiving a: "The model data archive is too large. Please reduce the size of the model data archive or move to an instance type with more memory." error. The tar.gz file is 7.7G but is not loaded in memory (only a small part of it is). I am wondering if anything changes recently that may be causing this issue.

Thanks for any insights

Emmanuel

1 Answer 1

1

When an Endpoint is created with SageMaker, the model data artifact is downloaded and uncompressed onto the associated disk/EBS volume on the instance. This volume size is proportional[1] to the instance type that you chose.

Please make sure that the instance type that you picked has enough disk space to accommodate the uncompressed .tar.gz file. (It does not matter if it is fully or partially loaded onto memory later on, it has to fit in the disk uncompressed).

[1] Volume size for instance types - https://docs.aws.amazon.com/sagemaker/latest/dg/host-instance-storage.html

Sign up to request clarification or add additional context in comments.

1 Comment

Is there any way to deploy models with huge size(around 60gb compressed)? Or should I go for multi-model endpoint in that case?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.