1
00:00:04,939 --> 00:00:11,060
In this lesson, we are going to talk about the challenges of the rack technique.

2
00:00:11,060 --> 00:00:20,900
Well, you have in the attached material a one document talking about the main challenges of the rack

3
00:00:20,900 --> 00:00:21,740
technique.

4
00:00:22,280 --> 00:00:30,050
A I don't think, uh, now is the best moment to go in detail about this.

5
00:00:30,050 --> 00:00:39,530
I just wanted to give you the idea that in the rack technique, like in almost anything, you know,

6
00:00:39,530 --> 00:00:48,380
in in programming, you have like a basic way of doing things, and then you have like an advanced way

7
00:00:48,380 --> 00:00:54,140
of doing this or to improve, you know, the basic functionality.

8
00:00:54,140 --> 00:00:56,810
So with the rack technique, it happens the same.

9
00:00:57,440 --> 00:01:06,140
You have like a basic way of doing things, and then you can start optimizing the different steps you

10
00:01:06,140 --> 00:01:07,190
are using.

11
00:01:07,190 --> 00:01:15,590
In the case of the rack technique, we have two main phases of the of the technique that what we call

12
00:01:15,590 --> 00:01:24,470
the retrieval phase when we are storing data, and what we call the response generation phase when we

13
00:01:24,470 --> 00:01:26,810
are searching for an answer.

14
00:01:26,810 --> 00:01:35,270
So in both of them you have ways to optimize the performance of the phase.

15
00:01:35,270 --> 00:01:35,720
Right.

16
00:01:35,720 --> 00:01:41,510
So we will see more about that uh, later in next, uh lessons.

17
00:01:41,510 --> 00:01:50,750
But it was important for us to tell you that we have some challenges in the, in the, in the, in the

18
00:01:50,750 --> 00:01:51,140
way.

19
00:01:51,140 --> 00:01:51,770
Right.

20
00:01:52,400 --> 00:02:03,680
So, so on every phase of the rack technique, we are able to use some optimization techniques to improve

21
00:02:03,680 --> 00:02:04,940
the performance.

22
00:02:04,940 --> 00:02:12,530
So we will see some techniques to overcome the challenges in each of the phases later.

23
00:02:14,480 --> 00:02:22,370
In the next lesson, we are going to talk about the main considerations to select the right orchestration

24
00:02:22,370 --> 00:02:24,500
framework in our opinion.

25
00:02:24,500 --> 00:02:35,180
So we are going to give us to give you the, the our honest opinion about a how to decide from long

26
00:02:35,180 --> 00:02:41,360
chain llama index or the API of open AI.