Lesson 3 In-Class Discussion

Part 1 (2018)
21

Which links?

22

I wonder if the blog links are found in a resource thread (my guess, not sure). Anyone know?

23

Blogs are discussed here

6 Likes
24

Disappointed that youtube chat is disabled today

25

CurlWget is here

5 Likes
26

I didn’t understand what is the benefit of using this chrome extension? Or what are the cases when we should use it?

27

It’s for when you want to download data in a shell when the data is hidden behind a password.

6 Likes
28

You often want to download data in a remote computer rather than in your own computer.

1 Like
29

If you have direct access to a file, you can just use wget (or rsync or several others). But if you need to be logged in, like for a DropBox link or something like that, the extension helps you get command-line access to it so you can download it to a server.

1 Like
30

Probably to keep discussions in the forum.

31

so bn_freeze() just applies batch normalization and freezes the layer right?

2 Likes
32

where is dog-cat quick?

1 Like
33

Yep :slight_smile:

34

is learner.bn_freeze(True) is equivalent to model.eval() in pytorch?

35

Where can we find the video recordings of the course?

36

It is not.

37

Batch normalization is applied on the forward pass anyway, because it’s a part of the model. “bn_freeze” freezes batch normalization layers to prevent updates to their parameters.

4 Likes
38

thanks.

39

I’m not seeing the Keras or the Quick Dog and Cat file, could someone link it or show the folders it’s in from the github. Just not seeing it . . .

40

You might have to pull the latest repository?