Can we schedule ml.t3.medium based sagemaker studio jobs?

0

The sagemaker job creation allows only ml.m4/m5 onwards instances. I have a simple requirement that can easily work with ml.t3.medium. Is there a way to schedule sagemaker studio jobs using ml.t3.medium? Thanks in advance.

  • SageMaker training jobs don't support t3.medium instances. Are you running notebooks or Python scripts? If it's simple enough, could you use Lambda or other serverless options?

demandé il y a 9 mois352 vues
2 réponses
0

@Durga_S Complete notebook. It can be refactored and made into python module probably. Compute wise it isn't big enough to require ml.m5.large. The cheapest route is probably through ec2 which is what I need to explore.

répondu il y a 9 mois
0

Unfortunately, there is no option to use ml.t3.medium to schedule a SageMaker Studio Job. The smallest instance type for this feature is ml.m5.large which you have already mentioned.

What are you hoping to achieve inside this notebook of yours? If it's to process some data then I would suggest trying out SageMaker processing job.

Processing job's intended purpose it to fetch raw data from S3 (or other sources), run the script that processes the data then sends the data back to S3. You'll be billed for only the duration of the job in seconds and there is an option to use ml.t3.medium as well.

Below is a code sample you can use as reference should you choose to use processing job: https://github.com/aws/amazon-sagemaker-examples/blob/main/sagemaker_processing/scikit_learn_data_processing_and_model_evaluation/scikit_learn_data_processing_and_model_evaluation.ipynb

AWS
BenH
répondu il y a 9 mois

Vous n'êtes pas connecté. Se connecter pour publier une réponse.

Une bonne réponse répond clairement à la question, contient des commentaires constructifs et encourage le développement professionnel de la personne qui pose la question.

Instructions pour répondre aux questions