WEBVTT 00:00.080 --> 00:00.600 Hey there. 00:00.640 --> 00:01.240 Eden here. 00:01.240 --> 00:05.760 And in this video we're going to be bootstrapping our environment for the project. 00:05.760 --> 00:08.800 So it's going to be a bit long around 15 minutes. 00:08.800 --> 00:14.560 And it's not going to be interesting because let's face it, bootstrapping a Python environment is not 00:14.560 --> 00:15.680 that interesting. 00:15.720 --> 00:17.840 We'll be cloning the course repository. 00:17.880 --> 00:20.720 We will create a new branch that will be working in. 00:20.760 --> 00:23.560 We'll be using UV as our package manager. 00:23.560 --> 00:27.880 We'll install some dependencies, configure the API keys for the LLM. 00:27.920 --> 00:31.960 We're going to be talking to defining environment variables. 00:31.960 --> 00:37.360 And by the end of this video we'll have our environment ready to write our first link chain chain. 00:37.720 --> 00:39.120 So let's go to the code. 00:39.120 --> 00:41.920 So this is the course's GitHub repository. 00:42.080 --> 00:45.120 And I'll be linking it in the video's resources. 00:45.120 --> 00:47.640 And you can also find it in the introduction section. 00:47.840 --> 00:52.360 And all the code that we'll be writing can be accessed via this repository. 00:52.360 --> 00:56.080 So I will be linking the exact commit code of every video. 00:56.080 --> 00:59.720 So you can have the exact code as I'm having in running. 00:59.760 --> 01:00.200 All right. 01:00.200 --> 01:02.520 So let's go now and set up our environment. 01:02.520 --> 01:05.880 And the first thing I want to do is to clone this repository. 01:06.160 --> 01:09.520 So I'm going to go and copy here this link here. 01:09.720 --> 01:11.560 And let me open my terminal. 01:12.920 --> 01:15.600 Let me go now and CD into my desktop. 01:15.600 --> 01:17.400 And here I'll be cloning the code. 01:18.360 --> 01:22.680 And let's now write git clone and paste here the URL. 01:25.360 --> 01:30.600 And now that it's done we can see into the directory which is called link chain course. 01:31.840 --> 01:34.680 And let me open my IDE from here. 01:34.680 --> 01:38.440 And I'm going to be using cursor for the next couple of videos. 01:38.440 --> 01:40.920 But you can use any IDE that you want. 01:41.320 --> 01:42.680 You can use PyCharm. 01:42.680 --> 01:44.240 You can use VSCode. 01:44.560 --> 01:47.200 Hey you can even use vim if you're really, really brave. 01:47.480 --> 01:49.960 The IDE that we'll be using doesn't really matter. 01:49.960 --> 01:52.640 Will be using basic functionality of Ides. 01:52.680 --> 01:55.040 So to run code to debug it. 01:55.040 --> 01:56.440 And that's pretty much it. 01:56.600 --> 02:03.130 And the funny fact is that I actually recorded this course when cursor did not exist, and I used PyCharm 02:03.130 --> 02:04.410 for most of the videos. 02:04.690 --> 02:05.170 All right. 02:05.210 --> 02:08.330 So let me open now the terminal from the IDE. 02:08.690 --> 02:13.930 And I want now to create a new branch which is going to be called Hello World branch. 02:14.090 --> 02:19.130 And this branch is going to hold all the code for the next couple of videos where we're going to introduce 02:19.130 --> 02:19.810 link chain. 02:19.970 --> 02:23.650 So I'm going to go and write git checkout dash dash orphan. 02:23.650 --> 02:27.490 And I'm going to call this branch project Slash Hello World. 02:28.210 --> 02:30.890 So when I do that I'm going to be in a new branch. 02:30.890 --> 02:33.490 You can see it right over here on the bottom left here. 02:33.770 --> 02:37.490 And the dash dash orphan creates a branch with no commit history. 02:37.490 --> 02:42.450 So it's going to start completely new without any commits from the existing repository. 02:42.490 --> 02:47.210 Now notice that by the time you'll be cloning the repository, this branch will already exist. 02:47.210 --> 02:49.770 So you might want to call this with a new name. 02:49.770 --> 02:50.210 Maybe. 02:50.250 --> 02:51.570 Hello world one. 02:52.250 --> 02:52.610 All right. 02:52.610 --> 03:01.570 Let me just write git dash rf from this directory to remove all the files that we have to really start 03:01.610 --> 03:03.090 with no files at all. 03:03.450 --> 03:05.490 So you can see we don't have any code. 03:05.490 --> 03:06.690 We don't have any commits. 03:06.730 --> 03:08.610 We are in a clean slate right now. 03:09.210 --> 03:09.690 All right. 03:09.690 --> 03:11.850 So let me write UV dash dash help. 03:12.010 --> 03:16.410 And UV is a very fast Python package manager just like pip. 03:16.410 --> 03:22.730 But it's much faster because it's built on rust and it handles installing and resolving and running 03:22.730 --> 03:24.770 dependencies very, very efficiently. 03:24.970 --> 03:29.610 And it's also going to help us manage our isolated environments with our dependencies. 03:29.850 --> 03:32.970 So we'll only be using the basic features of UV. 03:33.170 --> 03:37.050 And we'll be installing packages and managing our virtual environments. 03:37.250 --> 03:43.330 By the way, you can also use poetry or Pipenv or any other Python package manager you would like. 03:43.490 --> 03:48.490 And all we'll be doing in the course is installing packages and creating virtual environments, and 03:48.490 --> 03:50.370 running the code there with the packages. 03:50.490 --> 03:51.930 So nothing too fancy here. 03:51.930 --> 03:57.210 And in fact, we'll be using poetry and we'll be using BPMN through the entire course in some projects. 03:57.490 --> 03:58.250 All right. 03:58.250 --> 04:04.410 So let's go and enter UV Just to make sure that we have you've installed. 04:04.410 --> 04:06.930 Let me go up here in the logs. 04:07.170 --> 04:09.690 And yeah, you can see it's installed in my machine here. 04:09.690 --> 04:12.170 So we can see here the manual. 04:12.170 --> 04:18.130 And if you don't get this kind of output simply write Pip3 install UV and you should have it installed 04:18.130 --> 04:18.970 in your machine. 04:18.970 --> 04:21.850 So I already have it installed so it's not going to do anything. 04:22.370 --> 04:22.810 All right. 04:22.810 --> 04:25.810 So let's go now and initialize our project. 04:25.810 --> 04:28.170 And let's go and write UV init. 04:28.330 --> 04:31.570 And it's going to bootstrap a new Python project. 04:31.570 --> 04:34.730 And it's going to create some skeleton for our project. 04:34.730 --> 04:40.250 So it's going to create our main.py file with some boilerplate code, some hello world code. 04:40.490 --> 04:46.970 And it's going to create a Pyproject.toml file which is going to have the list of packages we have installed 04:47.290 --> 04:48.450 in our project. 04:48.810 --> 04:50.730 So right now this is a clean project. 04:50.730 --> 04:52.410 So we don't have anything installed. 04:52.570 --> 04:56.730 So let me go and write UV add and let's go and add link chain. 04:56.730 --> 05:00.050 And this is going to install link chain in our system. 05:00.330 --> 05:03.820 And notice that it also created here a folder. 05:03.820 --> 05:07.260 So it also now created the virtual environment for our project. 05:07.260 --> 05:10.700 And the package is now installed in this virtual environment. 05:10.700 --> 05:12.740 So let me go and click this play button here. 05:13.020 --> 05:16.260 And now notice that we're running the code here. 05:16.260 --> 05:19.180 So you can see here this print hello from LinkedIn course. 05:19.300 --> 05:24.900 And you can see that it's run from the virtual environment you see here LinkedIn course under those 05:24.900 --> 05:25.740 parentheses. 05:25.740 --> 05:28.900 So this means that it runs within this virtual environment. 05:29.180 --> 05:35.180 All right let me also add the link chain open AI package which is the link chain integration package 05:35.180 --> 05:36.180 for open AI. 05:36.540 --> 05:42.820 And it's a separate Python package that is going to contain all the open AI specific integrations for 05:42.820 --> 05:49.020 the link chain ecosystem and the reason why we are installing it independently, and it's not being 05:49.020 --> 05:56.820 bundled into the package itself, is because Link Chain decided to split providers, which are third 05:56.820 --> 06:03.420 party services, to link chain to different and independent packages where all the vendors maintain 06:03.420 --> 06:10.060 their own packages because we don't really want to download 100 providers if we only want to use OpenAI 06:10.180 --> 06:13.900 or anthropic or whatever service provider that we want to. 06:14.140 --> 06:21.020 Now, historically, the packages were actually bundled in the original package, but in one of their 06:21.020 --> 06:26.900 releases, they split everything into integration packages, which design wise makes a ton of sense, 06:26.900 --> 06:32.140 because right now every provider is going to maintain its own package. 06:32.140 --> 06:37.500 And if it's going to add new models or new features, it's going to maintain it independently from the 06:37.500 --> 06:38.540 other providers. 06:38.540 --> 06:40.980 So we have everything nicely decoupled. 06:41.460 --> 06:41.860 All right. 06:41.860 --> 06:43.060 Let me click enter. 06:43.420 --> 06:45.620 And now we have this package installed. 06:46.100 --> 06:53.500 Let me also add Python env which is going to help us load the environment variables from env files, 06:53.500 --> 06:58.340 which is going to be very easy to load all the API keys that we're going to be using in the course. 06:58.340 --> 07:01.900 And if you're not familiar with it, don't worry, I'll be showing it very, very shortly. 07:01.900 --> 07:09.180 Shorty and let me also install black, and I sought simply to format nicely the code here and there. 07:12.340 --> 07:13.820 All right, let me click enter. 07:14.140 --> 07:19.300 So we installed all the packages and let me show you the pyproject.toml file here. 07:19.660 --> 07:25.100 And we can see here all the packages that we have installed and all of their versions. 07:25.100 --> 07:27.140 And we have all the details here. 07:27.500 --> 07:33.540 And of course when you'll be running this code yourself, you're going to have different and newer versions 07:33.540 --> 07:34.900 of all those packages. 07:35.100 --> 07:41.140 And I will be updating the code and on GitHub and the videos themselves in case the API changes. 07:41.300 --> 07:44.940 But the code you should see should match the latest version of Linkchain. 07:45.060 --> 07:49.420 And if it's not the case, then please let me know and I will fix this ASAP. 07:49.460 --> 07:54.580 I refilmed this course like 3 or 4 times, so I'm really used to it. 07:54.900 --> 07:56.660 All right, let me create a. 07:57.020 --> 07:58.060 Gitignore file. 07:58.300 --> 08:04.070 Because there's a ton of things that we do not want to commit to the repository for example API keys 08:04.150 --> 08:06.830 or the Venv directory. 08:06.910 --> 08:09.470 So let me paste here a standard. 08:09.710 --> 08:11.910 Gitignore file for Python projects. 08:12.030 --> 08:17.430 So let me go now to my browser and let me look for Python. 08:17.470 --> 08:19.630 Gitignore a file. 08:19.670 --> 08:20.430 Let me go. 08:20.430 --> 08:23.150 And yeah let me check this project here. 08:23.750 --> 08:26.390 Let me go now and copy the content here. 08:26.790 --> 08:31.790 And by the way when you create a new GitHub project you have an option to generate these. 08:31.790 --> 08:33.390 Gitignore file automatically. 08:33.550 --> 08:38.790 But anyways let me just simply go and let me paste the content in my gitignore file. 08:39.670 --> 08:40.510 Alrighty. 08:40.910 --> 08:45.150 So now we can see that the venv directory is grayed out. 08:45.150 --> 08:46.750 So we're not going to commit it. 08:46.750 --> 08:48.510 And we're not going to track it with git. 08:48.910 --> 08:52.350 And now let's go and create a dot env file. 08:52.510 --> 08:57.870 And this file is going to hold all of our environment variables and all of our API keys. 08:58.110 --> 09:01.390 So now I want to fill in the OpenAI API key. 09:01.950 --> 09:07.710 And it's really, really important to note that you do not want to share your API keys. 09:07.910 --> 09:09.510 Do not share them with anyone. 09:09.510 --> 09:16.710 It's like a password and do not, of course, commit it to GitHub and the API keys, which I'm going 09:16.710 --> 09:21.270 to show you in the course, are all going to be revoked once I finish filming this video. 09:21.470 --> 09:27.470 So you do not want your API key to be leaked, because we have malicious people in the world which scan 09:27.510 --> 09:34.310 those kinds of files and those kinds of variables on GitHub automatically, and once they do find it, 09:34.310 --> 09:35.830 they go and abuse it. 09:35.830 --> 09:41.590 And you do not want to find yourself being charged with lots of money because your API key got leaked, 09:41.590 --> 09:44.670 or you accidentally committed it into a GitHub repository. 09:44.870 --> 09:48.750 And again, it's very important to note that I will be using OpenAI. 09:49.030 --> 09:55.150 I will also in the following video will show you how to use open source models with llama, and you 09:55.150 --> 09:56.950 can use any LLM that you want. 09:56.950 --> 09:58.190 You can use anthropic. 09:58.230 --> 09:59.510 You can use Gemini. 09:59.550 --> 10:01.550 You can use any LLM that you want. 10:01.750 --> 10:08.150 I also made a video, which is going to be in this section, which is going to tell for each project 10:08.150 --> 10:10.550 which LMS can be used for that project. 10:10.590 --> 10:11.030 All right. 10:11.030 --> 10:18.030 So let me search for OpenAI API key and let me go now to platform openai.com. 10:18.310 --> 10:22.630 Let me go and create an API key and let me call it link chain of course. 10:23.670 --> 10:28.150 And let me go and copy it and let me go and paste it. 10:28.510 --> 10:34.590 Now this is going to be working for me because I have already set up my OpenAI account and put credit 10:34.590 --> 10:40.670 card information, and I've already set up a budget that I'm going to limit my usage, which is also, 10:40.670 --> 10:45.350 by the way, a good practice to limit your usage and consumptions of APIs. 10:45.510 --> 10:50.750 So even if your API does get leaked, then you won't be maxing out the limit that you put. 10:51.150 --> 10:56.870 And a quick disclaimer OpenAI is not free, so you would need to purchase some credits beforehand. 10:57.030 --> 11:01.830 I will be showing you the free alternative to run open source models on our machines. 11:01.950 --> 11:05.400 Um, but this is going to be in the next video, not in this one. 11:05.600 --> 11:09.520 And in this video and in general in the course I will be using OpenAI. 11:09.760 --> 11:13.880 And you can use again any LLM that you want if you have free alternatives. 11:13.920 --> 11:19.720 Now it's important to note that if you're not going to have a credit card with OpenAI, then you will 11:19.720 --> 11:25.120 be encountering this 4 to 9 error, which is not going to let you use their API. 11:25.560 --> 11:30.840 And another thing is that the name here, OpenAI API key is very important. 11:30.840 --> 11:36.080 So the name should be the exact name, as you can see, because link chain is going to be looking for 11:36.080 --> 11:42.960 this exact environment variable when it's going to try making requests to the OpenAI API to get responses 11:42.960 --> 11:44.040 from their llms. 11:44.280 --> 11:51.120 By the way, if you want to use the Google API key, let me go and look for a Google AI studio. 11:54.000 --> 11:57.880 And here you can go and generate and create an API key. 11:58.720 --> 12:01.000 Let me just go and select a project. 12:01.000 --> 12:02.960 So it's going to be associated with it. 12:03.560 --> 12:06.480 and let me go and create now an API key. 12:06.720 --> 12:10.360 So this API key can be used for Google's models. 12:10.360 --> 12:11.320 So Gemini. 12:12.240 --> 12:19.280 So you need to write Google API key in case you want to use Gemini from the Google AI studio. 12:19.520 --> 12:24.560 And for that you'll need to install the link chain Google JNI integration. 12:24.560 --> 12:26.880 But I won't be showing that in the course. 12:27.200 --> 12:33.640 And in case you want to use open source models running from your local machine, you can install the 12:33.640 --> 12:35.920 link chain Dash Olama package. 12:36.040 --> 12:39.960 And I will be showing how to use link chain Olama in this section. 12:42.120 --> 12:49.360 All right let's go now and import from dot env the load dot env function. 12:50.040 --> 12:55.400 And when we call it it will go and look for a dot env file. 12:55.640 --> 12:59.200 And then it's going to take all the environment variables there. 12:59.200 --> 13:02.920 And it's going to load their values into our environment. 13:03.160 --> 13:10.520 So now after we run this, we can actually access those environment variables, the OpenAI API key that 13:10.520 --> 13:11.240 we configured. 13:11.640 --> 13:13.760 Let's see that this is indeed the case. 13:13.760 --> 13:18.800 So let me go and import the OS module so we can access the environment variable. 13:18.800 --> 13:24.880 And let me go and print here OS dot environment dot get. 13:24.960 --> 13:28.600 And here we want to get the OpenAI API key. 13:29.120 --> 13:30.520 So let's go and run it. 13:30.760 --> 13:34.000 And you can see now the value of my API key. 13:34.000 --> 13:35.840 And you should see your value here. 13:35.840 --> 13:40.840 So this means we can access the API key from our code and from our environment. 13:40.840 --> 13:43.280 And everything looks okey dokey right now. 13:43.280 --> 13:48.400 And we can continue and finally start writing our first link chain chain. 13:48.600 --> 13:50.680 So let me just remove this code from here. 13:51.200 --> 13:54.040 Let me run black to format the code nicely. 13:56.200 --> 13:59.520 Let me also run Eishort just in case. 14:00.160 --> 14:03.400 Alright, so now let's go and commit this code. 14:03.400 --> 14:04.890 So this boilerplate here. 14:05.290 --> 14:08.090 So let me go and add all of those files. 14:08.090 --> 14:09.410 Let me write git add. 14:11.450 --> 14:16.930 And now let's commit the code and let me call the commit environment setup. 14:19.250 --> 14:25.890 So let me push it and let me go copy this command and push it to the remote repository. 14:26.970 --> 14:27.490 All right. 14:27.490 --> 14:29.530 And let me show you where you can find the code. 14:30.570 --> 14:35.930 So you can go to the repo and look for the branch which is called oops. 14:35.970 --> 14:40.610 Let me refresh the page here and you can look for the branch which is called Hello World. 14:41.850 --> 14:44.970 And right over here let's go to the commit list over here. 14:45.410 --> 14:47.970 And here is the first commit environment setup. 14:47.970 --> 14:51.530 And here you can find all of the code that we wrote here. 14:51.930 --> 14:57.130 And of course this will be linked in the videos resource where you can have direct access to this link. 14:57.450 --> 14:58.290 And yeah. 14:58.290 --> 15:03.370 So now we can go and finally start writing our first link chain chain.