1
00:00:01,200 --> 00:00:06,750
Hello, guys, in this section, I will show you how you can create or how you can deploy your deep

2
00:00:06,750 --> 00:00:09,430
learning model on a decanted application.

3
00:00:09,870 --> 00:00:16,650
So maybe if you are creating a deep learning model for classification or for detection or something

4
00:00:16,650 --> 00:00:22,890
like that and you want to share it with your boss or with your friends, etc., so I think that it is

5
00:00:22,890 --> 00:00:30,450
a little bit ugly, if you will, Sharrett informative on a program with the model on a fine, etc..

6
00:00:30,690 --> 00:00:38,280
So your friends or your boss need need always to to run the program, to install the dependencies like

7
00:00:38,280 --> 00:00:45,570
the data flow, etc. to the library that you are using and all this stuff.

8
00:00:45,600 --> 00:00:47,550
Some it may take some time.

9
00:00:47,790 --> 00:00:52,470
So in this case, I think that this is a little bit great.

10
00:00:52,470 --> 00:00:58,770
If you will share your model or your example of deep learning as form as a teaching application, or

11
00:00:59,070 --> 00:01:06,420
you can create a website or that is a lot of times how a lot of types of how we can deploy your deep

12
00:01:06,420 --> 00:01:07,050
learning model.

13
00:01:07,890 --> 00:01:11,790
In this case, we will talk about how we can deploy it on a decanted application.

14
00:01:12,120 --> 00:01:16,650
We will not talk about the training part for the learning because it is not our subject.

15
00:01:16,980 --> 00:01:22,770
But although if you are interested, I will leave the link of the program for the training, for testing

16
00:01:22,770 --> 00:01:23,920
Sicherer in description.

17
00:01:24,180 --> 00:01:30,960
So in this case, you can you can take the you can go to the model and you can see if maybe if you are

18
00:01:30,960 --> 00:01:33,360
interested by this example so you can train in.

19
00:01:34,170 --> 00:01:39,990
But in our in our section for this section, we will talk only about how we can create this one and

20
00:01:39,990 --> 00:01:41,790
where we can put the model.

21
00:01:42,030 --> 00:01:49,710
Because if you will not put the or you will not put the line for loading the model in the right place,

22
00:01:49,950 --> 00:01:51,690
it will cause a problem, not a problem.

23
00:01:51,690 --> 00:01:54,170
But your application will be very, very slow.

24
00:01:54,390 --> 00:02:01,470
So if, for example, you will put the the model or the line that will load the model in this button,

25
00:02:01,470 --> 00:02:08,040
for example, on this button means that in each click on this button, you will look, we would reload

26
00:02:08,720 --> 00:02:09,340
the model.

27
00:02:09,690 --> 00:02:16,740
So in this case, we know all all we know that the when you will do the model, it will take a few seconds,

28
00:02:16,740 --> 00:02:19,050
maybe sometimes seconds, etc..

29
00:02:19,470 --> 00:02:24,410
So imagine with me on each click, you will you will wait for 30 seconds, 30 seconds.

30
00:02:24,690 --> 00:02:27,780
So it will not be great work.

31
00:02:27,790 --> 00:02:35,130
So if you will put the model in place when you will be load, only one time means that if we will open

32
00:02:35,130 --> 00:02:41,460
because here, if I will open this application, the model will be lowered only one time, then I don't

33
00:02:41,460 --> 00:02:44,520
need to load this another time because it is already loaded.

34
00:02:45,030 --> 00:02:48,110
So in this case, you will wait a few seconds at the beginning.

35
00:02:48,120 --> 00:02:50,220
After that you will use it without any problem.

36
00:02:50,670 --> 00:02:55,470
Then at the end of this section, I will show you how you can create this one, because this application

37
00:02:55,470 --> 00:02:59,640
is I didn't run it from this program because this is a program of application.

38
00:02:59,940 --> 00:03:03,120
But here I didn't run it because here I am.

39
00:03:03,120 --> 00:03:10,050
I created an executable file, which is an application that can be shared with your friends, with your

40
00:03:10,050 --> 00:03:10,950
boss, with anyone.

41
00:03:11,130 --> 00:03:19,200
And they they don't need to install any libraries, need that sort of loan or below for the images.

42
00:03:19,200 --> 00:03:22,230
If they will not, they won't need anything.

43
00:03:22,380 --> 00:03:26,130
They will need only this file, which is this application.

44
00:03:26,370 --> 00:03:33,360
So after clicking on it or running it, they will wait for a few seconds to load the the mother.

45
00:03:33,600 --> 00:03:37,050
After that, they can use it with with no problem like this.

46
00:03:37,050 --> 00:03:43,770
For example, here, if I click, I can't choose any type of birth as easily as I told you, because

47
00:03:44,670 --> 00:03:49,950
in this example my model is for image classification, which means for birth classification.

48
00:03:49,950 --> 00:03:53,580
I have two hundred and fifty five types of birds.

49
00:03:53,940 --> 00:03:59,670
So I will choose, for example, this one and you click here or predict and it will give you I think

50
00:03:59,670 --> 00:04:07,950
that you see here, I think that there is a problem with this one, which is another one here.

51
00:04:08,220 --> 00:04:08,670
Yes.

52
00:04:09,330 --> 00:04:15,000
Sometimes because I think that the model is not as robust, but this is not our our project.

53
00:04:15,270 --> 00:04:20,850
But here, as you can see here, it will give us the name of the the bird with the probability.

54
00:04:21,270 --> 00:04:28,230
But here, if I will choose, for example, another one like this one may be the same thing.

55
00:04:28,230 --> 00:04:34,740
It will give us the name of the or the bird with the probability of seeing this, because you will see

56
00:04:34,740 --> 00:04:37,170
if this bird is exactly the same or not.

57
00:04:37,180 --> 00:04:43,360
For example, if this will give us a small probability means that the model is not sure, etc..

58
00:04:43,770 --> 00:04:45,690
All of this stuff are for the deep learning.

59
00:04:45,690 --> 00:04:51,480
Our project is how we can create this application and how can we deploy the model for this application.

60
00:04:51,930 --> 00:04:58,510
So we will do this project in several videos because it will take time if we use all the stuff in the

61
00:04:58,520 --> 00:04:59,160
one video.

62
00:04:59,400 --> 00:05:00,720
So we divide them on.

63
00:05:00,800 --> 00:05:06,860
Song videos, how we can create this one and how we can put the modern or how we can deploy the modern,

64
00:05:06,870 --> 00:05:07,420
etc..

65
00:05:07,820 --> 00:05:10,490
So I see you in the next video by.