Free credits do not allow launching GPU-accelerated instances

Question

Free credits do not allow launching GPU-accelerated instances

Slavkó 0

Hi!
I have created a new account with $200 free credits to experiment with PyTorch models training in cloud, as described at https://learn.microsoft.com/en-us/azure/machine-learning/how-to-train-pytorch?view=azureml-api-2
After some "fun" with setting everything up, I've realized that:

My subscription's quotas allow only NC or NV GPU-equipped instance types
NV and NC instance types were discontinued in Sep 2023
I'm not allowed to request change to quotas while I'm on the "free credits" subsccription

My question is:
Is there a way to use free credits for running GPU VMs and run there basic training tasks?

VasaviLankipalle-MSFT 18,576 Reputation points

2024-03-14T00:03:04.8833333+00:00

Hello @Slavkó , Thanks for using Microsoft Q&A Platform.

From the document you have shared, under the pre-requisites, it's mentioned that.

before you can run the code in this article to create a GPU cluster, you'll need to request a quota increase for your workspace.

I can understand that you have issue with using instances based on the free credits. I would suggest check your usage and quotas under the Azure portal.

Also, as per the documentation, looks like only 750 hours each of B1s, B2pls v2 (Arm-based), and B2ats v2 (AMD-based) burstable VMs provided for student free account, unfortunately, this list does not include GPUs.

If you're using an Azure free account, you can use only a limited number of Azure compute cores.

I hope this helps.

Regards,
Vasavi
Slavkó 0 Reputation points

2024-03-14T11:58:56.27+00:00

Hi @VasaviLankipalle-MSFT and thank you for your response. I have checked my quotas before writing the initial question. They include only outdated GPU instances that cannot be started since Sep'23. Also, I'm not on a "student subscription". I am a new user who created an account hoping to experiment with Azure ML. Too bad it's not reflected anywhere that I cannot do this with free credits. Now I'm very frustrated because spent my time with no valuable result and will go back to Google colab, since it provides overall better experience and can provide me with T4 GPU for free at least for a limited amount of time.

Your answer

VasaviLankipalle-MSFT 18,576 Reputation points

2024-03-14T00:03:04.8833333+00:00

Hello @Slavkó , Thanks for using Microsoft Q&A Platform.

From the document you have shared, under the pre-requisites, it's mentioned that.

before you can run the code in this article to create a GPU cluster, you'll need to request a quota increase for your workspace.

I can understand that you have issue with using instances based on the free credits. I would suggest check your usage and quotas under the Azure portal.

Also, as per the documentation, looks like only 750 hours each of B1s, B2pls v2 (Arm-based), and B2ats v2 (AMD-based) burstable VMs provided for student free account, unfortunately, this list does not include GPUs.

If you're using an Azure free account, you can use only a limited number of Azure compute cores.

I hope this helps.

Regards,
Vasavi
Slavkó 0 Reputation points

2024-03-14T11:58:56.27+00:00

Hi @VasaviLankipalle-MSFT and thank you for your response. I have checked my quotas before writing the initial question. They include only outdated GPU instances that cannot be started since Sep'23. Also, I'm not on a "student subscription". I am a new user who created an account hoping to experiment with Azure ML. Too bad it's not reflected anywhere that I cannot do this with free credits. Now I'm very frustrated because spent my time with no valuable result and will go back to Google colab, since it provides overall better experience and can provide me with T4 GPU for free at least for a limited amount of time.

Share via

Free credits do not allow launching GPU-accelerated instances

Your answer