1 00:00:02,820 --> 00:00:09,930 What is long serve and why it is important for us to be aware of this new. 2 00:00:11,240 --> 00:00:12,050 A. 3 00:00:12,700 --> 00:00:16,059 Solution provided by the long chain team. 4 00:00:16,720 --> 00:00:26,020 So long, sir, is a one solution that is going to help us build an API for our LM application. 5 00:00:26,020 --> 00:00:30,910 And this is going to be a situation that we are going to find frequently. 6 00:00:30,910 --> 00:00:40,180 If you remember in a previous chapter, we have been talking about how to do this with a fast API directly, 7 00:00:40,330 --> 00:00:49,450 but the long chain team has been working on a um, on their own solution to do that, and they call 8 00:00:49,450 --> 00:00:51,730 it Lang Chef right now. 9 00:00:51,760 --> 00:01:03,940 Lang Chef Lang serve as a a November of 2023 is still in beta version, but, uh, we expect that it, 10 00:01:03,940 --> 00:01:12,670 uh, will uh, it will be better and better in order to, uh, provide more and more, uh, functionality 11 00:01:12,880 --> 00:01:21,970 also in production right now, we can, uh, access to Lang serve via Jupyter notebook, via the editor, 12 00:01:21,970 --> 00:01:28,720 like, uh, like Visual Studio Code or whatever editor you are using, and also via the terminal of 13 00:01:28,720 --> 00:01:29,740 your computer. 14 00:01:29,740 --> 00:01:37,420 In this exercise, we are going to provide a basic code for the Jupyter Notebook option. 15 00:01:37,420 --> 00:01:44,770 And you will see at the end of the exercise some links to the playground to test the the application 16 00:01:44,770 --> 00:01:47,350 and also the API routes, etc.. 17 00:01:47,350 --> 00:01:52,840 So in the right side of the screen you have the working goal for this. 18 00:01:52,840 --> 00:02:00,010 Initially, as always, we connect with the dot m file where we have the credentials to uh, be able 19 00:02:00,010 --> 00:02:02,800 to communicate with the API of OpenAI. 20 00:02:03,190 --> 00:02:09,370 And then you will need to install some, uh, packages in your environment. 21 00:02:09,370 --> 00:02:11,290 You will need to install Lang serve. 22 00:02:11,320 --> 00:02:18,640 You will need to install this other packages and you will need to install also a this modules. 23 00:02:18,640 --> 00:02:23,410 What we are doing here is we are creating a fast API application. 24 00:02:23,410 --> 00:02:29,320 We are creating a couple of routes with slash OpenAI and slash chain. 25 00:02:29,620 --> 00:02:34,930 And we are uh, creating the a local server. 26 00:02:34,930 --> 00:02:39,490 So you will see that in the local server route. 27 00:02:39,490 --> 00:02:42,310 We are going to have this content hello world. 28 00:02:42,310 --> 00:02:49,030 And that we are going to start the, uh, server in this, uh, location. 29 00:02:49,030 --> 00:02:57,370 So if you go to the a initial link to the, you will, you will see the playground to check the app. 30 00:02:57,370 --> 00:03:02,320 Then you have the API routes in slash OpenAI and slash chain. 31 00:03:02,320 --> 00:03:07,720 And finally you have the fast API, API documentation okay. 32 00:03:07,720 --> 00:03:12,280 So here's where you can start playing around with all these, uh, possibilities. 33 00:03:12,280 --> 00:03:21,280 So as you see, it's a very similar solution to the one we, uh, initially try with Fetch Fast API 34 00:03:21,280 --> 00:03:28,780 in a previous chapter, this is the option that Lang Chain is going to, uh, promote, uh, looking 35 00:03:28,780 --> 00:03:35,140 at the future, and we expect that it will be better and more complete and with more possibilities in 36 00:03:35,140 --> 00:03:35,890 the near future. 37 00:03:35,890 --> 00:03:43,120 So for you, it's important, in my opinion, to check the Lang Smith documentation about it and start 38 00:03:43,120 --> 00:03:47,950 preparing yourself for the moment where Lang server is out of beta. 39 00:03:47,950 --> 00:03:51,160 So it's it's like a solid solution.