Name: SAP Artificial Intelligence Onboarding : GitHub Setup
Uploaded: 2023-9-7
Duration: 784 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>In this video tutorial, Tahir Hussain Babar (Bob) shows how to onboard SAP Artificial Intelligence & how to setup pre-requisites. This video goes through setting up GitHub so that your Workflow files can be synchronized with AI Core. We also create both a Git Repository Connection and an Application in SAP AI Core Administration tab in AI Launchpad.<br><br>00:00 - Intro<br>01:03 - Git Repositories & Application in SAP AI Core Administration<br>02:40 - GitHub Repository Creation<br>03:39 - GitHub Application (Workflows Folder) Creation<br>04:52 - git_setup Creation<br>06:14 - GutHub Personal Access Token Creation<br>08:04 - Git Repository Creation in SAP Launchpad using Python<br>10:45 - Application Creation in SAP Launchpad using Python<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><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>

SAP Artificial Intelligence Onboarding : GitHub Setup

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 series we'll be looking at onboarding and in this video we'll be looking at how to connect your GitHub repository to your SAP AI Core Service. In the previous few videos, we've looked at using the SAP Business Technology platform to set up the SAP Artificial Intelligence Launchpad to connect to the SAP Artificial Intelligence Core Service. We then looked at setting up Jupiter so we can write commands to use that SAP Artificial Intelligence Core service using an API. In the previous video, again we looked at connecting to our workspace within the SAP AI Launchpad, but we didn't really get a result back to confirm that we can connect. So in this video, we're going to set up SAP AI Core administration by setting up connections to our GitHub repository. And then I'm going to show you the concept of creating an application which we can also connect to in GitHub. So why are we using GitHub? Is it used to store our source code? Well, the answer is actually no, we're not going to be using GitHub to do that. We're going to be using a GitHub repository to create what's called a GitHub repository application. And in that application, it's going to simply be a series of configuration files in YAML formats, which will be used when we're training and serving our machine learning models. And we'll look at these different workflow and serving templates later on in the video series when we're looking at machine learning and setting up our training and serving models. Now I've already taken a few seconds to create myself a GitHub account. That's all I've done. So if I go to Github and click on my link, all I've done is create a very basic Github account and I've got no repositories in it. If you don't know, well of course Github is also used for software development and version control using what's called Git. So you can of course, if you want to, store your code within a folder, within Github in a specific repository, but that's not the purpose of what I'm going to be recording in these videos. Again, it's to create those workflow templates in YAML format, which will be used later on in the video series. What might not be obvious is if you store your code within your GitHub repositories, of course don't store any of your secure keys or your service keys in your GitHub repository as it will be a security risk. So again I've done really no configuration here. All I've got is a very simple account. It literally took 20 seconds to create. There's no repositories. And if I go to the top right hand side and click on my profile, the user is sh8-ai-videos. And again you can see we have no public or private repositories. So what we'll do is we'll build ourselves a new repository. And I'll call this ai-core-test. And then I will go down, I can make it a private repository, we can choose to add a readme, and then at the bottom I'll click on create repository. So after a few seconds that repository will be created and again our workflow template files will be stored in a folder and we're going to call that folder workflows. So what we'll need to do here is have a subfolder within the repository called workflows. Now you can't create a folder here without having an object in it. But there's kind of a workaround here within our repository which we can do. If i create myself a new file, we'll put in the name of the folder that we want to create called workflows and then we'll do a forward slash x and that will essentially create an empty file within a folder called workflows. Today I'm going to be talking about how to create a workflow template for your YAML application. We've got our folder, we don't have our YAML workflow templates yet, that's going to be covered later on in the video series. But what we want to do is back in our AI Launchpad, we'll want to connect our Git repository and create what's called that application, which is essentially linked to that Workflows folder. We can do this on the right hand side with the Add button. We can't really do this here with the application, so we've got the repository. But I'm actually going to show you how to do this, of course, using the SAP AI SDK. So back in our Jupyter Notebook, what I'll do is I'll save the file. What we'll need to do first is create another connection file to our GitHub repository. Now the form of the file will look like this. So you'll see there are two sections for repo and app on line 8 and line 2. So the first thing we might want to do is save the file. So I'm going to rename this file as git underscore setup dot json. I'll select rename and I'll just save this file so we've got it saved in our files folder. So you can see on line 4 and on line 10 we're requested to put in our GitHub user. So again here it's SHA-AI-videos. So I'll go back to line 4 and I'll change where it says write your GitHub username as that name which is SHA-AI-videos like so. Now in terms of line 3 we need to give the repo a name. It's a good idea not to name it the same as your URL, so I'm going to call it myai-core-test-as-one-word. We can also use that command that's in line 4 for line 5 because it's the same thing. The username is going to be shai-videos as well as course for line 10. It's the same thing here for the app. Now for the GitHub password it's not a good idea here to place in your actual GitHub password. What you should use is something called the personal access token. So I'll show you how to set that up. So if I go back to my profile in Gitsub and go to settings, if I scroll right to the bottom we go to developer settings and at the bottom you've got personal access tokens right at the bottom. So what I'll do is I'll select it and I'll generate a new token. You'll be prompted to log in with the password for your Github account. And then you just need to specify what you're using the token for. So for example, I'm only going to use it for a few days, for seven days. And then you need to decide on the rights that you want. So for us, it will be mainly to know the repo and workflow status. So when you've added your rights, all you need to do then is click on generate token. And this is the password that you'll use. Of course, if you want, you can also revoke all or delete this specific personal access token as well. So I deleted that personal access token and created another one that lasts a couple of weeks. And I can paste that here. If we look at line 12, you'll see that there's a path which is workflows, which of course indicates that folder that we created in GitHub. And you've got your application name on line 9, which I'm going to rename to be My AI Core Test App. And again, the workflows path here on line 12, that relates to what we've got in GitHub, which is this folder that we created as well, which will contain our workflow templates in YAML format. So now that we've done this, all we need to do is specify file save to save the file. And we can close the file. So now that we've got that connection file, of course, we can go back to our notebook and execute some commands. So like before, what we want to do is read that git underscore setup.json file. So we'll set it as a variable. Then, of course, we want to run the command to actually read the file and store it. So here what we're doing is we're loading the github setup.json file and we're connecting the various parameters from that file to our repository such as the username, the URL, the name of the repo and of course the password. The last two lines simply print the response. So what I'll do is I'll execute this cell, and the response indicates that the repository has been onboarded. That's the output message. To be actually sure there are other commands that we can execute, for example this one, will simply get a list of all the repositories that are available to us. So if I execute this, as you can imagine at the moment we've got one repository called myai-core-test. And of course to be absolutely sure if we go back to AI Launchpad and refresh, within git repositories we can see we've got that myai-core-test git repository created. Now once your Git repository has been created, you can edit and remove the connection. But of course you could do this back in your Jupyter Notebook as well. So here's a sample command which will delete this specific repository. So if I run it, we can see that the repository has been removed from the tenants. Of course if we go back and refresh now that git repository is gone. All I would now need to do is rerun the original command to onboard the repository, get the list of repositories as we've just seen here, and of course the ultimate test is to go back and refresh then we can see that the git repository has reappeared. As I said before it's a good idea to create a header for each of these commands, again it's like self-documentation. So what I can do is I can create another cell, and then we can call this, if I move the cell with the up and down keys and we change it to heading, I can change the heading to as follows. So we can call it link repository to SAP AI Core. That's what we're doing here. So it's a good way of documenting your code. Of course, it's another idea to write comments with the gate key, with the gate character as well in your code. So after saving the file, what we might want to do as well is create our application. So if I copy this, I'll paste this command below. Of course, we need to make sure that it's actually a heading like so. And I'll change repository to app. Again, what we'll do is we'll add another cell where we can paste in our code. So back in AI Launchpad, you'll see we've got our repository, but we haven't created an application yet. Remember, this application will be the YAML files in that workflows folder. We can create the actual application here, but of course I'm going to create it in the Jupyter Notebook. So the command has the same structure as you've seen before. We'll load the GitHub setup.json file, and then we'll register the directory within that GitHub setup file as an app, which again will be that workflows folder. When we run, we've actually not printed that we want a response, but again we've got those other commands where we can see how many applications we've got. So here we can see we've got the application pointing to this repository. The application is called myai-core-test-app and again remember this app will contain your YAML files for training and serving your models in that workflows folder. So all we need to do is check. What we'll actually do is go back to the app and we can see here we've got the application as well of course as our GitHub repository. To summarize, we've linked our GitHub repository to AI Core and of course we've linked our application to AI Core as well. In the next two videos we'll be looking at setting up and connecting Docker to AI Core service. Again, a Docker repository needs to be connected to SAP AI Core for developing, shipping and running applications within containers. And these containers will be used to both train and serve your models.