1 00:00:03,290 --> 00:00:07,850 Okay, so this is probably the production level. 2 00:00:07,850 --> 00:00:17,780 Uh, LM application, the professional, uh, LM application that is making a more sophisticated use 3 00:00:17,780 --> 00:00:20,570 of, uh, the Smith platform. 4 00:00:20,570 --> 00:00:24,470 This is the chat dot Lang Chain.com project. 5 00:00:24,470 --> 00:00:30,080 This is a support chat created by the Lang Chain team. 6 00:00:30,290 --> 00:00:36,080 And here in this notebook, you will see the second notebook. 7 00:00:36,110 --> 00:00:46,310 You will see in the in the in this blog you have a link to a video, uh, where uh, they explain in 8 00:00:46,310 --> 00:00:51,290 detail how this professional project is using Lang Smith. 9 00:00:51,290 --> 00:00:54,860 I will highlight a few screens of that video. 10 00:00:54,860 --> 00:01:02,960 But the interesting thing about this project is that we have the Lang Chain team, uh, explaining us 11 00:01:02,960 --> 00:01:04,610 how they use Lang Smith. 12 00:01:04,610 --> 00:01:11,150 And you will see they serve very interesting information, but they also share the code of this project. 13 00:01:11,150 --> 00:01:17,840 So here in the notebook you have also a link to the GitHub repository, where there they share the code 14 00:01:17,840 --> 00:01:18,620 of this project. 15 00:01:18,620 --> 00:01:25,850 And within the code of this project you can see also sophisticated code related to Lang Smith. 16 00:01:25,850 --> 00:01:29,810 So some of the screens I have selected for you. 17 00:01:29,810 --> 00:01:34,910 So this is a video you can find in the YouTube channel of Lang Chain. 18 00:01:34,910 --> 00:01:40,190 And in February 2024 a the uh. 19 00:01:41,560 --> 00:01:49,720 The person responsible from the team in in in long chain for the the project long Smith, which is one 20 00:01:49,720 --> 00:01:56,980 of the co-founders of long chain is explaining us in this video is like a one hour video. 21 00:01:57,460 --> 00:02:01,840 How do they use Lang Smith in the in the project? 22 00:02:01,840 --> 00:02:07,810 And this is a very interesting project because as you can see here in this screen, this is a project 23 00:02:07,810 --> 00:02:12,850 that is using like, you know, hundreds of thousands of tokens. 24 00:02:12,850 --> 00:02:21,820 They are spending a significant amount of money, uh, in, in, in, in open AI, you know, to have 25 00:02:21,820 --> 00:02:27,310 this project running, they are using different models, but in open AI they are they are making an 26 00:02:27,310 --> 00:02:28,660 interesting investment. 27 00:02:28,660 --> 00:02:35,710 And this is probably the project that is making the best use of Lang Smith, because this is a project 28 00:02:35,710 --> 00:02:39,130 developed by the the, the Lang Smith developers. 29 00:02:39,130 --> 00:02:39,430 Right. 30 00:02:39,430 --> 00:02:48,010 So so it's very interesting to see, you know, under the hood what they are uh, doing with Lang Smith. 31 00:02:48,010 --> 00:02:53,710 So this is one of the screens that I have selected for you. 32 00:02:53,710 --> 00:02:59,290 I have selected just a few screens just to tell you what you are going to find when you see the video. 33 00:02:59,290 --> 00:03:01,180 I recommend you to see this video. 34 00:03:01,180 --> 00:03:05,620 To study this video in detail, you will see things like. 35 00:03:05,770 --> 00:03:11,740 If you remember, we talk about the metadata and the different metadata possibilities you can use in 36 00:03:11,740 --> 00:03:15,880 order to monitor and and make a B, testing, etc.. 37 00:03:15,880 --> 00:03:19,480 So this is another screen that is going to be very interesting for you. 38 00:03:19,480 --> 00:03:21,370 Look at the metadata. 39 00:03:21,370 --> 00:03:30,250 You can you can see in this project also here you have uh in the monitor dashboard how they are comparing 40 00:03:30,250 --> 00:03:33,400 different versions of the of the project. 41 00:03:33,400 --> 00:03:40,210 So very interesting for you to take a look at this and also take a look at the right side bar where 42 00:03:40,210 --> 00:03:47,200 they are sharing some of the some of the different tags and filtering options they are working with. 43 00:03:47,200 --> 00:03:50,470 You have here the different test. 44 00:03:50,530 --> 00:03:56,290 Uh, do you remember that in our basic example we only use one test in the data set. 45 00:03:56,290 --> 00:04:00,520 So here you have different tests and the comparisons they are doing. 46 00:04:00,520 --> 00:04:06,340 And finally here you have also a more detailed comparison between different versions. 47 00:04:06,340 --> 00:04:06,940 Okay. 48 00:04:06,940 --> 00:04:14,950 So I just wanted to tell you that we are very lucky to have uh, this video from the Long Chain team. 49 00:04:15,040 --> 00:04:20,680 Uh, I don't know how to call it the long chain team or the long Smith team. 50 00:04:20,680 --> 00:04:27,730 You know, they are the same company, but, uh, anyway, Long Chain or Lang Smith has shared this 51 00:04:27,730 --> 00:04:28,930 video with us. 52 00:04:28,930 --> 00:04:35,110 They also have shared the code of the application with us in GitHub. 53 00:04:35,110 --> 00:04:42,520 And I think this is an invaluable resource for us because the creators of Lang Smith are showing us 54 00:04:42,520 --> 00:04:46,240 how they are using Lang Smith for their own projects. 55 00:04:46,240 --> 00:04:48,640 So very, very interesting project. 56 00:04:48,640 --> 00:04:56,440 I, I really recommend you to go here, uh, and, and check this video, study this video in detail 57 00:04:56,440 --> 00:05:04,270 because they are sharing a very, uh, a number of very interesting, uh, uh, subjects with us and 58 00:05:04,270 --> 00:05:07,240 also recommendations and uh, advices. 59 00:05:07,240 --> 00:05:09,160 So very, very interesting. 60 00:05:09,160 --> 00:05:13,210 In my opinion, this is right now the best source of information. 61 00:05:13,210 --> 00:05:17,500 You want to see how a professional project is using. 62 00:05:17,500 --> 00:05:18,940 Uh, Lang Smith today.