Request for assistance with training yolov8 model in Sagemaker Studio and changing GPU instance

Question

Hey,I am new to AWS and I want to train my yolov8 model in sagemaker studio. The disk size is quite small in the that. My dataset is of arround 5 gb. Can someone please tell me how i can i do that and how can I change the GPU instance?

i also would like to that if i connect S3 to sagemaker studio, will i be able to bigger disk space?

Any help will be greatly appreciated.

i also would like to that if i connect S3 to sagemaker studio, will i be able to bigger disk space?

Any help will be greatly appreciated.

Answer

Hi, if you're using Studio Classic, you can change the instance type on the fly from the notebook, see instructions here - https://docs.aws.amazon.com/sagemaker/latest/dg/notebooks-run-and-manage-switch-instance-type.html

The Studio storage uses Elastic File System (EFS), and expands as you use, so you should be able to use the 5GB dataset. If you run into out of memory errors in your notebook, try going up an instance size. 
Connecting to S3 does not affect your storage size in Studio.

For the [new Studio experience](https://docs.aws.amazon.com/sagemaker/latest/dg/studio-updated.html), you can stop the JupyterLab space and increase the EBS storage associated to the space within the Studio UI. The maximum EBS storage supported is 16384 as per the docs.

Request for assistance with training yolov8 model in Sagemaker Studio and changing GPU instance

Conteúdo relevante