aws-samples/sagemaker-ml-workflow-with-apache-airflow

This repository shows a sample example to build, manage and orchestrate Machine Learning workflows using Amazon Sagemaker and Apache Airflow.

sagemaker-ml-workflow-with-apache-airflow's Issues

DAG Import Errors

Hello, I just set up airflow environment using Airflow 2.0.2 RDS CloudFormation.
When I access and login the airflow UI page, I got a above error.

Do you know what is problem?

Thanks,
gwangjin

Content of container

I see here that you are using a container to build a custom model.

Would it be possible to have more details on how this container is structured and understand the input/output of the estimator for different channels (e.g. train, tune, etc.)?

sagemaker version - get_image_uri() not resolved

I am trying to replicate your way of defining the DAG for a sagemaker training job. But, for this line:

from sagemaker.amazon.amazon_estimator import get_image_uri

I get this error:

Cannot find reference 'get_image_uri' in 'amazon_estimator.py' .

Is it defined in any specific version? Or would you have any suggestions for resolving this error, please?

Thanks

Airflow not getting deployed on EC2

Hi,

After running the CloudFormation script, it looks like the airflow install is broken. Can you please take a look?

"ImportError: No module named json"
user-data.log

user-data.log attached.

Thanks!

Sagemaker error when deployed on EC2

Hi guys,

I install sagemaker==v1.40.0

After I launch CloudFormation, I get error on EC2 log
ClientError: An error occurred (ValidationException) when calling the CreateModel operation: 2 validation errors detected: Value '{{ ti.xcom_pull(task_ids='model_training')['Training']['TrainingJobName'] }}' at 'modelName' failed to satisfy constraint: Member must have length less than or equal to 63; Value '{{ ti.xcom_pull(task_ids='model_training')['Training']['TrainingJobName'] }}' at 'modelName' failed to satisfy constraint: Member must satisfy regular expression pattern: ^a-zA-Z0-9*

Note: I checked with sagemaker==v1.39.2 ==> it worked