1
00:00:10,840 --> 00:00:11,050
Okay.

2
00:00:11,050 --> 00:00:16,420
So in this lecture, I want to give you guys a little bit of a disclaimer before we proceed to the installation

3
00:00:16,420 --> 00:00:17,200
lectures.

4
00:00:17,200 --> 00:00:22,870
So the thing you have to keep in mind is that these installation lectures are only guidelines.

5
00:00:23,110 --> 00:00:23,280
Okay?

6
00:00:23,290 --> 00:00:28,480
These are generic lectures that I created at one point in time to cover all my courses at once.

7
00:00:28,480 --> 00:00:33,640
And obviously this is the most scalable way to distribute such lectures because I don't want to have

8
00:00:33,640 --> 00:00:39,250
to create another lecture installing stuff from scratch every single time I create a course, which

9
00:00:39,250 --> 00:00:40,420
is pretty often.

10
00:00:40,840 --> 00:00:47,920
So just to give you guys a little bit of history about why this lecture exists, in fact, this lecture

11
00:00:47,920 --> 00:00:48,940
never used to exist.

12
00:00:48,940 --> 00:00:55,180
So when I first created my courses, I didn't have any installation lectures because I didn't know that

13
00:00:55,180 --> 00:00:57,640
people didn't know how to install Python.

14
00:00:57,640 --> 00:01:01,660
And this is because Python was a prerequisite to those courses.

15
00:01:01,660 --> 00:01:07,270
So basically if you came to the course and you didn't know how to install Python or install Python libraries,

16
00:01:07,300 --> 00:01:09,880
you basically failed to meet the prerequisites.

17
00:01:09,880 --> 00:01:16,490
Now, because I'm such a nice person, I decided to include an installation lecture eventually to handle

18
00:01:16,490 --> 00:01:17,540
these cases.

19
00:01:17,540 --> 00:01:22,790
So I was honestly surprised that people trying to get into machine learning didn't know how to install

20
00:01:22,790 --> 00:01:23,540
Python.

21
00:01:23,720 --> 00:01:28,190
Okay, this is kind of like learning Python today and then trying to learn machine learning tomorrow,

22
00:01:28,190 --> 00:01:28,580
right?

23
00:01:28,580 --> 00:01:31,850
That's a that's a very compressed way of learning machine learning.

24
00:01:31,850 --> 00:01:34,670
And that's not generally how it works in the real world.

25
00:01:34,730 --> 00:01:34,850
Okay.

26
00:01:34,910 --> 00:01:40,250
So in the real world, how it's usually going to happen is you're going to learn Python first.

27
00:01:40,250 --> 00:01:46,940
So you're going to have so you're going to have a pretty good handle on Python syntax writing basic

28
00:01:46,940 --> 00:01:51,710
programs in Python, maybe even not basic programs like more complicated programs.

29
00:01:51,710 --> 00:01:55,880
But in general, you know, installing Python is like the first step of that.

30
00:01:55,880 --> 00:02:00,650
So if you don't know how to install Python and you don't know how to use Python, how are you even going

31
00:02:00,650 --> 00:02:05,960
to implement a complicated machine learning algorithm in Python, you know, two days after learning

32
00:02:05,960 --> 00:02:06,680
Python?

33
00:02:06,710 --> 00:02:08,450
I don't think that's realistic.

34
00:02:09,800 --> 00:02:14,390
So one of the things you have to keep in mind is my famous rule, which applies to a lot of things in

35
00:02:14,390 --> 00:02:18,350
this course, is learn the principles, not the syntax.

36
00:02:18,830 --> 00:02:19,040
Okay?

37
00:02:19,040 --> 00:02:25,010
So what this means is it's more about understanding what we are doing and why we are doing it.

38
00:02:25,010 --> 00:02:25,190
Okay?

39
00:02:25,190 --> 00:02:26,780
It's about understanding.

40
00:02:27,230 --> 00:02:32,540
Whereas if you're focused solely on the syntax, okay, what this means is that you're probably taking

41
00:02:32,540 --> 00:02:36,080
a more monkey see monkey do approach, which is not correct.

42
00:02:36,110 --> 00:02:39,560
It's not about just typing exactly character for character.

43
00:02:39,560 --> 00:02:42,350
What I typed, it's more about understanding.

44
00:02:42,350 --> 00:02:47,690
So for example, I show you how to install a few libraries using PIP, so it's pip install, blah blah

45
00:02:47,690 --> 00:02:48,260
blah.

46
00:02:48,410 --> 00:02:52,550
So this doesn't mean you're going to install every library I installed.

47
00:02:52,580 --> 00:02:58,790
This means I'm showing you how to install a library so that if you come across a library that you need

48
00:02:58,790 --> 00:03:02,330
that you haven't yet installed, now you know how to install it.

49
00:03:02,360 --> 00:03:02,540
Okay.

50
00:03:02,570 --> 00:03:08,330
So that's sort of understanding the theme of what we are doing rather than just rote copying of the

51
00:03:08,330 --> 00:03:12,320
characters that I'm typing on the screen, which doesn't require any understanding at all.

52
00:03:12,560 --> 00:03:12,770
Okay.

53
00:03:12,800 --> 00:03:14,930
So it's understanding that's important.

54
00:03:14,930 --> 00:03:16,790
So if you're understanding, that's good.

55
00:03:16,790 --> 00:03:19,250
And if you're not understanding, that's not good.

56
00:03:20,000 --> 00:03:20,200
Okay.

57
00:03:20,210 --> 00:03:25,010
So as an example of this, in these lectures, you're going to see me install some libraries called

58
00:03:25,010 --> 00:03:29,420
Cntk and Theano, both of which are no longer maintained.

59
00:03:29,420 --> 00:03:35,270
So Theano was the first deep learning library to take advantage of the GPU in Python, which was the

60
00:03:35,270 --> 00:03:37,670
most popular around maybe ten years ago.

61
00:03:37,670 --> 00:03:41,420
And then we have S.a.k, which was Microsoft's Deep Learning library.

62
00:03:42,370 --> 00:03:43,600
So don't install these.

63
00:03:43,630 --> 00:03:45,160
These are just examples.

64
00:03:45,280 --> 00:03:50,350
Another example is OpenAI Gym, which is used in some courses, but not all courses.

65
00:03:50,350 --> 00:03:52,930
So you really want to pay attention to what course you're in.

66
00:03:52,960 --> 00:03:58,900
For example, if you're in a reinforcement learning course where we are using OpenAI gym, then that's

67
00:03:58,900 --> 00:04:00,490
a case where you'd want to install it.

68
00:04:00,490 --> 00:04:05,950
If you're not in a reinforcement learning course and you have no idea what OpenAI Gym is, then don't

69
00:04:05,950 --> 00:04:06,820
install it.