Algorithm arn does not exist when trying to start training job

Question

When trying to start a training job in sagemaker using AlgorithmEstimator (by inputting the algorithm arn), I get an error saying that the Algorithm arn does not exist.  I have tried this with multiple algorithm arns (both obtained from my own algorithms and algorithms from AWS Marketplace) and they all give a similar error message.

Here is a sample notebook that I am trying to execute but fails: [https://github.com/aws/amazon-sagemaker-examples/blob/main/aws_marketplace/using_algorithms/autogluon/autogluon_tabular_marketplace.ipynb](https://github.com/aws/amazon-sagemaker-examples/blob/main/aws_marketplace/using_algorithms/autogluon/autogluon_tabular_marketplace.ipynb)

I am using sagemaker==2.145.0, both locally and in a sagemaker notebook instance.

Here is a snippet of the error log:
```
ClientError: An error occurred (ValidationException) when calling the CreateTrainingJob operation: Algorithm arn:aws:sagemaker:us-east-2:[acc-id]:algorithm/arn:aws:sagemaker:us-east-2:[acc-id]:algorithm/autogluon-tabular-v3-5-cb7001bd0e8243b50adc3338deb44a48 does not exist.
```

Answer

Hi - It seems in that example the ARN is coming from https://github.com/aws/amazon-sagemaker-examples/blob/main/aws_marketplace/using_algorithms/autogluon/src/algorithm_arns.py and should be in the format e.g for us-west-2 it should be something like this **arn:aws:sagemaker:us-west-2:594846645681:algorithm/autogluon-tabular-v3-5-cb7001bd0e8243b50adc3338deb44a48**

It seems from your example that format is not correct. Can you please check what is the output of algorithm_arn you get when you run the cell from notebook autogluon_tabular_marketplace.ipynb
```
# Specify algorithm ARN for AutoGluon-Tabular from AWS Marketplace.  However, for this notebook, the algorithm ARN
# has been specified in src/algorithm_arns.py file and you do not need to specify the same explicitly.

from src.algorithm_arns import AlgorithmArnProvider

algorithm_arn = AlgorithmArnProvider.get_algorithm_arn(region)

```

Algorithm arn does not exist when trying to start training job

相關內容