How to run a tensorflow program on an EC2 instance while staying in the free tier

I've attempted to setup an EC2 instance so I can run my Tensorflow script. However, I've noticed that I cannot setup the instance with a storage space of < 75 GiB.

Tensorflow

< 75

GiB

This is the error message I get...

Launch Failed

Volume of size 30GB is smaller than snapshot 'snap-030808799cdf9332b', expect size >= 75GB

Everything is within the free tier range except the storage space. It is only free tier for the storage space if it is set to < 30 GiB.

< 30

GiB

I am running the following AMI which is free tier...

AMI

Deep Learning AMI (Ubuntu) Version 12.0 - ami-d1c9cdae


My instance type is t2.micro which is also free tier.

t2.micro

Everything is free tier except for my storage space due to the space needed for my snapshot. I cannot change the snapshot so I am forced to go outside of the free tier.

Is there any work around to this? Is there anyway I can run Tensorflow without having to pay for it?

Tensorflow

1 Answer
1

AWS Snapshots cannot be restored to an EBS volume smaller then the snapshot. You can restore to a larger EBS volume.

Your option is to use a standard Linux AMI such as Amazon Linux 2, Ubuntu, etc. and then install Tensorflow yourself. If you are learning Tensorflow, I recommend going thru the process of installation and configuration.

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.