1
00:00:03,920 --> 00:00:13,280
So in this detailed video we are going to see very interesting advanced tips regarding the Lamb-smith

2
00:00:13,280 --> 00:00:14,750
platform, for example.

3
00:00:16,370 --> 00:00:23,750
Deploying multiple versions of the LM app to production and monitor performance.

4
00:00:24,430 --> 00:00:25,300
You can.

5
00:00:25,300 --> 00:00:26,470
So how to do that?

6
00:00:26,470 --> 00:00:32,680
How to deploy multiple versions of your application to production and monitor the performance of these

7
00:00:32,680 --> 00:00:33,970
multiple versions.

8
00:00:34,630 --> 00:00:42,010
So you can route percentages of the traffic to different versions of your application in production.

9
00:00:42,930 --> 00:00:52,770
Or you can route all the traffic to just one production app and clone that traffic in a shadow pipeline

10
00:00:52,770 --> 00:00:54,450
with a different version.

11
00:00:55,460 --> 00:00:56,720
Super interesting.

12
00:00:57,020 --> 00:01:01,850
Okay, so just one quick tip about that.

13
00:01:01,970 --> 00:01:07,760
Second, what LM models can you use with Lang Smith?

14
00:01:08,480 --> 00:01:15,740
In principle, the Lang Smith team tells us that you can work with any LM model.

15
00:01:17,270 --> 00:01:26,870
Third tip, can you use Lang Smith if your LM application is not developed using Lang Chain again?

16
00:01:27,230 --> 00:01:28,160
Yes.

17
00:01:28,910 --> 00:01:39,380
You have different ways to use Lang Smith via the API and via the SDK, Python or TypeScript.

18
00:01:39,410 --> 00:01:47,330
Just a reminder, we said this before, but I think it's interesting to remark this to highlight this.

19
00:01:48,880 --> 00:01:55,000
Next, can you use Lang Smith with a multimodal application?

20
00:01:55,000 --> 00:01:55,840
This is.

21
00:01:56,020 --> 00:01:57,550
This is super interesting.

22
00:01:57,550 --> 00:02:04,870
And, uh, we are you will see that we are going to include, uh, a section about this later.

23
00:02:05,440 --> 00:02:11,260
Uh, so can you use Lang Smith with multimodal applications?

24
00:02:11,740 --> 00:02:15,370
Lang Smith is starting to experiment with this.

25
00:02:15,400 --> 00:02:19,690
It currently works with ChatGPT for vision.

26
00:02:20,080 --> 00:02:28,780
Super interesting because I bet we are going to see a lot of multimodal applications in the coming months.

27
00:02:28,780 --> 00:02:34,480
So it's very good to know that Lang Smith is preparing for that.

28
00:02:34,480 --> 00:02:39,220
And finally, how to deploy your project to Lang Serve.

29
00:02:39,670 --> 00:02:49,330
Just a reminder, this option is in beta and only available for plus, plus and enterprise versions.

30
00:02:49,330 --> 00:02:58,420
Okay, so you can do it is still in beta and it is only available for the Plus and Enterprise versions.

31
00:02:58,420 --> 00:03:05,320
Remember that we are going to see this in detail in the next lesson, but a Lang Smith currently has

32
00:03:05,320 --> 00:03:09,550
three versions the developer, the plus, and the enterprise.

33
00:03:09,550 --> 00:03:11,320
The developer is the free version.

34
00:03:11,980 --> 00:03:20,050
And this option, uh, the one click deployment in Lang Serve is just available for the two higher versions,

35
00:03:20,050 --> 00:03:21,460
plus an enterprise.

36
00:03:21,460 --> 00:03:21,970
Okay.

37
00:03:21,970 --> 00:03:31,810
And it's still in beta, but the Lang Smith team tell us that a they are going to invest in this, uh,

38
00:03:31,810 --> 00:03:32,680
feature.

39
00:03:32,680 --> 00:03:38,410
And I, I expect to have it ready and better very soon.

40
00:03:38,410 --> 00:03:39,010
Okay.

41
00:03:39,010 --> 00:03:48,010
In the next in the next lesson we are going to talk about some guides and recommendations that the Lang

42
00:03:48,010 --> 00:03:53,770
Smith team has shared with us and I think are very interesting to know.

43
00:03:53,770 --> 00:03:57,220
So let's see that in the next lesson.