Name: SAP AI ; Defect Detection ; Connecting S3 to AI Core
Uploaded: 2023-9-7
Duration: 485 s

Share feedback

<p><u><strong>Planned maintenance of Practice Systems</strong></u></p>

Planned maintenance of Practice Systems

Why create an SAP account?

Enhance your learning experience

SAP Learning

Choose your subscription

Manage your subscription

Live Sessions

Hands-on Practice

Certification

My Learning

Ready to start learning?

<div><div><div><div><div><div><div>In this playlist by Tahir Hussain Babar (Bob), from the SAP HANA Academy, we use SAP Artificial Intelligence to build a Deep Learning Image Segmentation Model which will help us perform automated Defect Detection on images of products. In this video, we revisiting using our AWS S3 Bucket, and then look at creating dataset artifacts in ML Operations in SAP AI Launchpad. <br><br>00:00 - Intro<br>01:07 - Revisiting AWS CLI<br>03:06 - Connecting S3 to AI Core<br>04:29 - Creating the S3 Service Key<br><br>SAP AI in BTP: <a href="https://www.sap.com/products/artificial-intelligence.html" rel="nofollow noopener noreferrer" target="_blank">https://www.sap.com/products/artificial-intelligence.html</a><br><br>Please note:<br>                     This video was originally published on the SAP HANA Academy YouTube Channel                      and therefore may contain references to SAP HANA Academy and YouTube.                      The tech world is always evolving and the SAP HANA Academy YouTube channel                      is set to close on December 31, 2023.                      At that time, references will no longer be valid                      and will gradually be removed when the videos are updated.<br>Also, this video includes automatically generated captions                      and therefore might misrepresent the spoken content or display incorrect product names.</div></div></div></div></div></div></div>

SAP AI ; Defect Detection ; Connecting S3 to AI Core

Welcome to the SAP HANA Academy. My name is Bob, and in this video series, we'll be looking at SAP Artificial Intelligence. In this video playlist we'll be building a deep learning model on product images for automatic defect detection using SAP AI Core. In this video we'll be linking the connection within our S3 bucket to SAP AI Core. In the previous video we looked at some of the differences between objects that can be shared in resource groups and those which can't. One of the things that can be put in a resource group is a dataset and that's one of the things that we're going to have to create when we try and train our model. So in terms of other objects that can be shared they are configurations, executions, deployments, datasets and models. These can go into resource groups. So we're not really going to build a data set here, but we're going to prep our connection to S3 for our data set. So what I'll do is I'll go to a command prompt, and what we'll just do is review our S3 bucket. What we'll do is we'll run the following command which is an AWS space configure. We've already done this a few videos ago. Here is our AWS access key ID which is already there as well as our password, our default region name and our output format. So we've already set this up earlier on in the onboarding series and earlier on in this series as well when we uploaded our data set to our S3 bucket. To show the data we're going to run an aws s3 ls and the name of our bucket command, because we want to look at the data that we've got which will of course be passed to our docker image. So if I put a forward slash image at the end forward slash we should have the data folder. And then within that we have our folders not good, not good mask, okay and okay mask as we've looked at in the first video. So these are of course going to be used by those docker images when we're training the model. We've got no good data and we've got good data which we've called OK. Again, just to confirm, we might want to look at all the data, and we can use a recursive command to list out all the contents, but I might want to look at all the data in the NG folder. So all we'll need to do is supply a forward slash ng at the end of this and we should see a list of all our no good images. So we know our S3 bucket works and we can connect and we have data and again we loaded that in the first video in the series. So what we'll need to do is configure connectivity to our S3 bucket from SAP AI Core. So we'll start off with creating our typical heading. So I'm going to call this, I'll make it a heading statement, like so. And we'll say that we're connecting S3 to AI Core. Now to do this of course we're using an object store which has a service key. So it's the same concept as when we connected to docker and github and also our ai core service. We're going to use the service key that we created when we created the instance. We covered that of course in the onboarding series within the last video. So if I go and paste this command, this is the key that we're going to have to create in the files folder. It'll be called S3 underscore service underscore key. Within files obviously we don't have it, we haven't connected to an S3 bucket yet. So we're going to do that here from scratch. So I'll go back to my instance, and if you remember when we created the object store, alongside the instance you'll have your service key. So for the instance for the object store on the right hand side, within the three dots, I can actually select it. Today we're going to look at how to create a new service key. If I click on the three dots, I can view that service key. So all I actually have to do is copy the contents from this file, and place it in that filename s3-service-key.json. So we'll go back to Jupyter, and we'll create that empty file. So I'll create myself a new file, and we'll paste in that text. And then I don't need to change anything here at all. We can leave all the content in the file. All I have to do of course is give the file a name. So I'm going to call this one S3 underscore service underscore key dot json. And we can save and we can close this file. So the file exists, so now we can run the commands we need to in that Jupyter Notebook. So I'll go back to the Notebook. Again, like normal, we're loading the service key and we're assigning it to a variable S3ServiceKeyPath. We'll then need to load that service key. And then lastly, the most important part is the connection name and all the parameters for that connection. So if I paste in what it looks like, it's just basically reading mostly from that service key apart from the path prefix. Hello, welcome to my lecture. And what we're creating is something called an object store secret, which is kind of like a Docker registry secret. It's a key. And we also know the connection parameters based on this default.secret. So all I need to do is actually run this command. In this command, we can see that the object store secret has been created. And now we can run a command which identifies which object stores are in a specific resource group. So in our resource group which is called image. So it's going to get all of those object store secrets from that resource group. So I'll execute that. And here you can see we've actually just got one connection to that S3 bucket as you can see here. Now you might be thinking, do I have a data set now in my list of machine learning operations in the Launchpad? Well the answer is no. All we've done is link our S3 bucket to AI Core. So I'll save this script and what I'll do is I'll go back to the Launchpad. Here we can see we've got no data set, but maybe it's because we've selected the wrong resource group. So I'll select the other one default and we'll go back to data sets, it's empty. And we'll go to hours and go back to data sets and it's still empty. In the next video, I'm going to explain why and how we need to build configurations now to link everything together before we can train our model. And we'll cover that in the next video.