Closed bimanata closed 7 months ago
Thank you very much for your feedback. We will investigate this matter and provide you with feedback.
This issue is stale because it has been open for 14 days with no activity.
This issue was closed because it has been inactive for 14 days since being marked as stale.
Describe the issue
I need detailed documentation regarding user and role management provided in the Sagemaker Extension configuration
Currently I found the user and role management is exciting capability, since we plan to use it on organization structure But still unsure how the role and policy provided segregate between access
My current setup look like this
Admin Access
-> train:all, checkpoint:all, inference:all, sagemaker_endpoint:all, user:all, role:all The expectation of this role is to administrate user and role, as well providing endpoint, common general model, checkpoints, datasets, that will be used by the other roleTrainer role
-> sagemaker_endpoint:all, train:all, checkpoint:all, inference:all The expectation of this role is to perform training of models, this role can upload their own model and ckpt but also can use the common general model that provided by the admin access.But looking at current setup, seems that any model that uploaded by
Admin access
not able to be seen by theTrainer role
. But whatever Trainer role upload, the Admin access seems able to see and use it. Can the team help what the role need to be setup for this kind of situation? And also if the role and policy can be detailed in the documentation and common practice for role management could be helpful as wellLinks
https://github.com/awslabs/stable-diffusion-aws-extension/blob/main/README.md