1 00:00:00,330 --> 00:00:03,030 So in the last section, you've made your API. 2 00:00:03,480 --> 00:00:09,900 No, that's true, that API into an actual full webpage that someone or anyone in the world with internet 3 00:00:09,900 --> 00:00:11,680 access can use. 4 00:00:11,730 --> 00:00:14,910 So let's open the book 75 and let's see how we do that. 5 00:00:15,600 --> 00:00:17,010 So let's begin here. 6 00:00:17,430 --> 00:00:19,590 We're opening that notebook. 7 00:00:20,820 --> 00:00:23,280 So let's plug in my repo here. 8 00:00:24,610 --> 00:00:29,880 Install flask and grub install. 9 00:00:30,690 --> 00:00:32,040 Navigate to the directory. 10 00:00:32,040 --> 00:00:35,580 Download the file and install it. 11 00:00:42,550 --> 00:00:47,680 Next, we have my what could there that is, Steve, that he did in the previous lesson. 12 00:00:49,150 --> 00:00:53,440 So now let's take a look at this file so you can see it looks quite simple. 13 00:00:53,590 --> 00:00:55,180 This is all the code there is. 14 00:00:55,270 --> 00:00:56,350 So what is this doing? 15 00:00:57,010 --> 00:01:05,860 So basically, we have in the GetUp report clue and we have the files that do the common functions here 16 00:01:05,860 --> 00:01:10,270 that do to do computer vision parts that you can take a look at these if you want. 17 00:01:10,750 --> 00:01:16,330 It's basically this one that gets to model, transforms the image and then formats the class name as 18 00:01:16,330 --> 00:01:18,220 well so that it returns it to the user. 19 00:01:19,540 --> 00:01:25,450 And we also have the inference function here, which is what I was looking for. 20 00:01:25,930 --> 00:01:32,200 This is one that takes to gets a prediction out of it from the transform and get model classes sort 21 00:01:32,200 --> 00:01:32,890 of functions. 22 00:01:33,550 --> 00:01:37,660 So basically this we have all the functions that we need here. 23 00:01:37,670 --> 00:01:43,770 We have flash, we have in grok, we have post requests and also can get requested. 24 00:01:43,780 --> 00:01:45,430 But let's look at the post requests here. 25 00:01:46,030 --> 00:01:52,740 So it actually loads a little website and that website here is index HTML so indexed. 26 00:01:52,750 --> 00:02:01,120 Each HTML is live in Templates directory and that basically consists of the HTML site. 27 00:02:01,120 --> 00:02:02,680 We can take a look at that here. 28 00:02:03,340 --> 00:02:08,050 This is the HTML of the webpage that we'll be deploying to the web. 29 00:02:08,380 --> 00:02:11,230 Once we start this app up, so enough talk. 30 00:02:11,230 --> 00:02:13,060 Let's see this and maybe can help you. 31 00:02:13,900 --> 00:02:16,780 This can help you understand better what we're doing. 32 00:02:16,990 --> 00:02:22,540 So let's click this link here and you're going to get or I didn't get to Google warning me because I 33 00:02:22,540 --> 00:02:23,560 opened it up before. 34 00:02:24,100 --> 00:02:29,250 So now we can upload an image here, so let's take a look. 35 00:02:29,260 --> 00:02:30,430 The screenshots? 36 00:02:32,260 --> 00:02:35,770 Let's go to my downloads, which is quite messy right now. 37 00:02:37,720 --> 00:02:40,570 What would be a good image to use? 38 00:02:42,010 --> 00:02:44,080 Actually can try a random QR code. 39 00:02:44,800 --> 00:02:45,920 Let's see what happens. 40 00:02:46,030 --> 00:02:50,050 This was a QR code I was doing some test with and it says Packet. 41 00:02:50,590 --> 00:02:52,380 So not that cool. 42 00:02:52,390 --> 00:02:53,620 Let's try something else. 43 00:02:53,950 --> 00:03:01,690 So, OK, so let's try this very nice split of a bowl of ramen soup. 44 00:03:02,230 --> 00:03:03,940 Let's see what our classifier thinks. 45 00:03:03,940 --> 00:03:05,060 It is remote. 46 00:03:05,100 --> 00:03:06,610 It is trained on image that. 47 00:03:06,730 --> 00:03:09,310 So let's see, it sets pot pot pipe. 48 00:03:10,450 --> 00:03:11,110 It's kind of cool. 49 00:03:11,170 --> 00:03:11,860 Let's try. 50 00:03:11,860 --> 00:03:16,390 Another image from my satellite was Try this chihuahua. 51 00:03:16,780 --> 00:03:18,010 Well, let's see what happens. 52 00:03:19,090 --> 00:03:19,630 Jhalawar. 53 00:03:19,660 --> 00:03:20,430 So we've got it right. 54 00:03:20,440 --> 00:03:24,400 So you can see our little web up here is working quite well. 55 00:03:24,640 --> 00:03:26,770 Let's try some fireworks. 56 00:03:26,980 --> 00:03:28,780 Yeah, let's see if it has fireworks. 57 00:03:28,780 --> 00:03:37,660 As a class surgeon, it did not have fireworks in the class, so I'm guessing maybe fireworks can look 58 00:03:37,660 --> 00:03:39,460 like a sea and perhaps actually can. 59 00:03:39,820 --> 00:03:41,800 It's kind of cool that it could, actually. 60 00:03:42,490 --> 00:03:46,480 Let's take a look at this dog here. 61 00:03:46,640 --> 00:03:48,670 Rottweiler, if he gets that one. 62 00:03:49,690 --> 00:03:51,160 And you can see it's working quite well. 63 00:03:51,700 --> 00:03:54,760 So that's it for this section here. 64 00:03:54,790 --> 00:03:55,750 Hope you enjoyed it. 65 00:03:56,110 --> 00:04:02,590 And Gronk and Flask make it very easy to test these web apps in the cloud so you can experiment here 66 00:04:02,590 --> 00:04:04,390 where you can experiment on your local machine. 67 00:04:04,870 --> 00:04:06,730 Would running this on your local machine? 68 00:04:06,730 --> 00:04:12,820 You don't need an grok anymore, but however you do if you want to actually host it and a web if you 69 00:04:12,820 --> 00:04:17,560 don't, you can just access it to your local host address here, and it's going to be something similar 70 00:04:17,560 --> 00:04:18,730 to this as well. 71 00:04:19,720 --> 00:04:25,570 So thank you for watching, and we'll stop there for now, and I'll see you in the next section. 72 00:04:25,660 --> 00:04:26,080 Thank you.