1 00:00:00,240 --> 00:00:00,640 Okay. 2 00:00:00,640 --> 00:00:04,200 This is the single most important moment of the entire lab. 3 00:00:04,680 --> 00:00:09,320 So it's the chat function, and I've spaced it out a bit to really explain what's going on. 4 00:00:09,320 --> 00:00:14,080 It's a more powerful chat function that includes the calling of a tool. 5 00:00:14,560 --> 00:00:17,120 It starts exactly the same way as it did before. 6 00:00:17,240 --> 00:00:19,120 We create some messages. 7 00:00:19,120 --> 00:00:20,640 Role of system. 8 00:00:20,640 --> 00:00:22,440 The content is the system prompt. 9 00:00:22,560 --> 00:00:24,280 We add in the history from before. 10 00:00:24,680 --> 00:00:25,680 Role of user. 11 00:00:25,680 --> 00:00:27,680 The content is the user's message. 12 00:00:28,200 --> 00:00:35,120 And now do a pretty standard trick of looping again and again and again until a variable called done 13 00:00:35,160 --> 00:00:36,520 gets set to true. 14 00:00:36,680 --> 00:00:39,480 So it's going to keep going until done is true. 15 00:00:39,880 --> 00:00:41,160 What is it going to keep doing? 16 00:00:41,600 --> 00:00:43,360 It's going to call open AI. 17 00:00:43,880 --> 00:00:46,640 And this is the usual thing OpenAI. 18 00:00:49,280 --> 00:00:51,360 Create we're familiar with that. 19 00:00:51,400 --> 00:00:52,680 We pass in the model. 20 00:00:52,920 --> 00:00:56,320 We pass in the messages this set of of messages. 21 00:00:56,560 --> 00:00:59,520 But look we also pass in Tools. 22 00:00:59,720 --> 00:01:06,080 Tools is that JSON blob, that structure of JSON, which is going to be sent to the LM saying, these 23 00:01:06,080 --> 00:01:07,360 are all the things you can do. 24 00:01:07,360 --> 00:01:09,760 This is the description of everything you can do. 25 00:01:10,280 --> 00:01:10,680 All right. 26 00:01:10,720 --> 00:01:12,800 And back it comes into response. 27 00:01:12,800 --> 00:01:18,880 And what comes in response might just be the output to go to the user, or it might be a response that 28 00:01:18,880 --> 00:01:21,600 indicates that tools need to be called. 29 00:01:21,600 --> 00:01:23,520 So that is what we do next. 30 00:01:23,560 --> 00:01:25,880 We look at response choices. 31 00:01:28,800 --> 00:01:35,200 And this tells us whether the LM is finished and it's done or whether it wants to call a tool. 32 00:01:35,640 --> 00:01:40,080 So if the finish reason is tool calls, it wants to call a tool. 33 00:01:40,480 --> 00:01:46,960 Then we we look at the message itself, we pluck out the tool calls from it. 34 00:01:47,560 --> 00:01:53,880 And this this is where we call that function that we just wrote, that clever function, that sneaky 35 00:01:53,880 --> 00:01:54,600 function. 36 00:01:54,840 --> 00:01:55,760 Right here it is. 37 00:01:55,760 --> 00:01:57,560 Remember it handle tool calls. 38 00:01:57,650 --> 00:02:02,810 This is the one that is the glorified if statement, the one that has the fancy if statement. 39 00:02:02,810 --> 00:02:07,970 But there's also this more silly version here that just has an actual if statement. 40 00:02:08,210 --> 00:02:09,850 But it's going to call this. 41 00:02:10,210 --> 00:02:11,490 So back we go. 42 00:02:12,170 --> 00:02:16,850 So again if it wants to call tools then we do that. 43 00:02:16,850 --> 00:02:17,930 We handle the tool calls. 44 00:02:17,930 --> 00:02:19,010 We call the tools. 45 00:02:19,170 --> 00:02:21,570 And then into the set of messages. 46 00:02:21,570 --> 00:02:25,130 We put what it wanted to do and the results of doing it. 47 00:02:25,490 --> 00:02:29,890 And then we will loop back and call a second time. 48 00:02:29,890 --> 00:02:32,810 And we will keep doing this until the finish. 49 00:02:32,810 --> 00:02:34,770 Reason is not tool calls. 50 00:02:34,770 --> 00:02:40,850 It does not want to call tools with whatever, it's just responded, in which case we return the content 51 00:02:40,850 --> 00:02:45,250 because that is going to be the actual response to the user. 52 00:02:45,610 --> 00:02:46,930 I hope that made sense. 53 00:02:46,970 --> 00:02:50,450 If not, go through it slowly, carefully until it does. 54 00:02:50,650 --> 00:02:52,330 This is this is pivotal. 55 00:02:52,330 --> 00:02:54,690 This is how the whole thing hangs together. 56 00:02:55,170 --> 00:02:56,450 Now, I know what you're thinking. 57 00:02:56,450 --> 00:02:58,650 You're thinking it's all very well showing us all this code. 58 00:02:58,650 --> 00:03:00,170 But does it actually work? 59 00:03:00,410 --> 00:03:01,770 Well, let's find out. 60 00:03:01,770 --> 00:03:03,170 Let's run the code. 61 00:03:03,210 --> 00:03:04,370 Here we have it. 62 00:03:04,370 --> 00:03:05,450 Let's say hi there. 63 00:03:05,810 --> 00:03:06,170 Hello. 64 00:03:06,170 --> 00:03:07,370 How can I assist you today? 65 00:03:07,370 --> 00:03:08,250 What's your job? 66 00:03:09,490 --> 00:03:11,650 I'm currently the co-founder and CTO of Nebula. 67 00:03:11,690 --> 00:03:12,570 Very nice. 68 00:03:12,730 --> 00:03:14,050 Uh, do you have a patent? 69 00:03:15,490 --> 00:03:16,850 Yes I do. 70 00:03:16,890 --> 00:03:17,650 Very nice. 71 00:03:17,730 --> 00:03:21,410 Uh, who is your favorite musician? 72 00:03:22,210 --> 00:03:23,210 That's what I want to know. 73 00:03:25,090 --> 00:03:29,370 I hope you, uh, heard that success. 74 00:03:29,490 --> 00:03:30,050 So it says. 75 00:03:30,050 --> 00:03:32,010 I'm not sure about my favorite musician. 76 00:03:32,010 --> 00:03:33,130 I appreciate a variety. 77 00:03:33,130 --> 00:03:35,010 If you have a favorite, I'd love to hear about it. 78 00:03:35,050 --> 00:03:36,370 It called the tool. 79 00:03:36,610 --> 00:03:37,570 You heard my phone. 80 00:03:37,570 --> 00:03:38,010 Go. 81 00:03:38,210 --> 00:03:39,890 We have success. 82 00:03:39,970 --> 00:03:42,810 The agent was able to take that JSON. 83 00:03:42,850 --> 00:03:45,570 It knew that this was a tool that it could call. 84 00:03:45,610 --> 00:03:48,410 It was able to respond with the finish. 85 00:03:48,410 --> 00:03:50,250 Reason was was tool calls. 86 00:03:50,250 --> 00:03:56,420 And that caused us to to interpret that to make the call to R our function handle tool calls, which 87 00:03:56,420 --> 00:04:03,500 had our fancy code that ended up calling this function that ended up calling pushover, that then sent 88 00:04:03,540 --> 00:04:07,820 the push notification to my phone and it all hung together. 89 00:04:08,180 --> 00:04:09,660 All right, let's keep going. 90 00:04:09,980 --> 00:04:11,220 I'd like to get in touch. 91 00:04:12,220 --> 00:04:13,020 That's great to hear. 92 00:04:13,020 --> 00:04:14,260 Please share your email address. 93 00:04:14,300 --> 00:04:19,340 I'm at Edward donna.com. 94 00:04:23,340 --> 00:04:24,740 I know you're expecting it. 95 00:04:24,740 --> 00:04:25,100 Now. 96 00:04:25,100 --> 00:04:27,380 The, uh, the magic's worn off. 97 00:04:27,420 --> 00:04:28,780 You're not impressed anymore? 98 00:04:28,900 --> 00:04:30,500 Well, I think that's really cool. 99 00:04:30,620 --> 00:04:36,220 So it has indeed, of course, notified me about my own email address, which is great. 100 00:04:36,220 --> 00:04:39,260 And so it shows you that it does hold together. 101 00:04:39,420 --> 00:04:44,460 Uh, and you are indeed able to have these interactions and get it to text you. 102 00:04:45,060 --> 00:04:45,780 Okay. 103 00:04:46,100 --> 00:04:46,860 That's great. 104 00:04:46,860 --> 00:04:48,420 I hope you enjoyed it. 105 00:04:48,420 --> 00:04:54,900 But next it's all about doing this for you and actually deploying this live to production so that you 106 00:04:54,900 --> 00:04:59,180 can serve your own avatar on your personal website. 107 00:05:00,060 --> 00:05:02,940 So if you've followed everything so far, then many congratulations. 108 00:05:02,940 --> 00:05:07,660 If you haven't, then also congratulations because it gives you this great opportunity to go through 109 00:05:07,820 --> 00:05:09,500 work through this until you do. 110 00:05:09,540 --> 00:05:10,700 And if I can help. 111 00:05:10,740 --> 00:05:13,260 Email me or contact me anytime. 112 00:05:13,260 --> 00:05:13,860 LinkedIn with me. 113 00:05:13,860 --> 00:05:16,980 Message me so that I can help you out. 114 00:05:17,300 --> 00:05:24,900 So what I want to do now is show you how you can deploy this application in production for yourself, 115 00:05:24,900 --> 00:05:28,100 so that you can have this as your virtual resume. 116 00:05:28,380 --> 00:05:30,460 Surely this is the future of resumes. 117 00:05:30,500 --> 00:05:36,900 No longer will we have profiles or CVS resumes where you list out your skills and experience, but rather 118 00:05:36,900 --> 00:05:41,220 you'll have a chat bot that people can interact with to learn about your career. 119 00:05:41,220 --> 00:05:48,020 And what better way to highlight your your AI abilities and your abilities to work with Agentic AI than 120 00:05:48,020 --> 00:05:55,870 to have an agentic solution up on your website that will allow you to interact with people and talk 121 00:05:55,870 --> 00:05:56,910 about your career. 122 00:05:57,510 --> 00:06:01,550 So let me show you some steps to deploying this. 123 00:06:01,950 --> 00:06:06,590 So the first thing to do is to put the code into a Python module. 124 00:06:06,590 --> 00:06:09,910 And I have done that right here for you in App.py. 125 00:06:10,230 --> 00:06:13,710 And you'll see that this has all of the stuff you would expect. 126 00:06:13,710 --> 00:06:20,550 It's got the same information in here, but organized into nice little functions at the top of the functions 127 00:06:20,550 --> 00:06:22,430 that cover the tools that we call. 128 00:06:22,750 --> 00:06:26,710 And then there is a class me as in me or or you. 129 00:06:27,070 --> 00:06:33,070 And that's something which has the handle tool call that we know so well that's right here. 130 00:06:33,070 --> 00:06:40,110 And it has the system prompt and this loop that we just went through, this chat loop that has the while 131 00:06:40,150 --> 00:06:46,590 not done it calls open I create, it passes in the tools. 132 00:06:46,710 --> 00:06:48,670 I don't think I mentioned this actually, so I'm pleased. 133 00:06:48,670 --> 00:06:51,430 I pause for a minute here because I do want to show this one more time. 134 00:06:51,630 --> 00:06:53,590 We call chat completions create. 135 00:06:53,630 --> 00:07:00,270 We select the model, we pass in the messages, and we pass in the tools JSON, the description of the 136 00:07:00,270 --> 00:07:01,550 tools that it can call. 137 00:07:01,910 --> 00:07:04,790 And with what comes back, we take choices zero. 138 00:07:04,910 --> 00:07:08,030 And we look at the finish reason back in the response. 139 00:07:08,030 --> 00:07:11,230 And we see whether or not that is tool calls. 140 00:07:11,430 --> 00:07:19,510 And if so, then we loop through to make sure that we run all of the tools before calling it again to 141 00:07:19,550 --> 00:07:20,950 get the final response. 142 00:07:21,350 --> 00:07:26,910 So this is so critical, this this chat with its loop that I do encourage you to take a look at it. 143 00:07:27,390 --> 00:07:30,950 And now at the bottom here we have the Gradio code. 144 00:07:30,950 --> 00:07:38,270 We've got it set up here so that if this is just run as it is then it will create a gradio chat interface 145 00:07:38,470 --> 00:07:44,550 with showing me chat and with the messages and it will launch it. 146 00:07:44,950 --> 00:07:49,840 So what I can do now, in fact, is I can bring up a terminal which you may remember. 147 00:07:49,840 --> 00:07:53,200 You press control and the backwards tick mark. 148 00:07:53,360 --> 00:07:56,200 I can go into the week one folder. 149 00:07:56,200 --> 00:07:59,200 Here we are and I can now type. 150 00:07:59,240 --> 00:08:01,120 Remember you don't type Python. 151 00:08:01,160 --> 00:08:04,040 You type, you run to run something. 152 00:08:04,040 --> 00:08:06,960 In this way you've run app.py. 153 00:08:07,680 --> 00:08:11,720 And if I run that then it should come up. 154 00:08:11,960 --> 00:08:13,520 So this link. 155 00:08:13,520 --> 00:08:17,920 Now on this link I can press control and click it to bring it up. 156 00:08:18,400 --> 00:08:19,160 Up it comes. 157 00:08:19,200 --> 00:08:20,840 Let me make that a bit bigger for you. 158 00:08:21,120 --> 00:08:26,240 This is now an app running on my box and I can say hi there. 159 00:08:27,040 --> 00:08:27,400 Hello. 160 00:08:27,400 --> 00:08:34,560 How can I assist you and, and uh, do you have a patent and so on and there we go. 161 00:08:34,760 --> 00:08:35,360 Okay. 162 00:08:35,680 --> 00:08:43,240 So so that works going back here, what we now want to do is look at actually deploying this. 163 00:08:43,560 --> 00:08:48,920 So there are various ways you can deploy an application like this, but one of them that's particularly 164 00:08:48,920 --> 00:08:55,800 simple and elegant is brought to you by the lovely AI company Hugging Face, which also happens to be 165 00:08:55,800 --> 00:08:57,960 the company that owns Gradio. 166 00:08:58,000 --> 00:09:02,960 That beautiful, framework hugging face has something called hugging face spaces, and hugging face 167 00:09:02,960 --> 00:09:06,880 spaces is a way that you can deploy things really simply. 168 00:09:06,880 --> 00:09:10,280 And then if you wish, you can also embed them in your own home page. 169 00:09:10,280 --> 00:09:13,680 So it gives you a really nice, simple way to do it. 170 00:09:13,960 --> 00:09:15,880 So this is how you do it. 171 00:09:15,880 --> 00:09:21,400 You first have to go to Hugging Faces website and set up your own account if you don't already have 172 00:09:21,400 --> 00:09:21,880 one. 173 00:09:21,880 --> 00:09:25,360 So you may already have a hugging face account, but if not, you need to set one up. 174 00:09:25,760 --> 00:09:31,640 Once you've done with that, you go into the foundations folder in a terminal and you run this command 175 00:09:31,680 --> 00:09:33,400 Gradio deploy. 176 00:09:34,000 --> 00:09:39,240 It's then going to ask you a series of questions, a bunch of simple questions to which I've got the 177 00:09:39,240 --> 00:09:40,920 answers you have to give right here. 178 00:09:41,360 --> 00:09:42,760 And I'll go through and do it right now. 179 00:09:42,760 --> 00:09:46,370 But I'm not going to do it properly, because one of the questions it's going to ask you about some 180 00:09:46,370 --> 00:09:52,890 secrets that you'll say yes, and you're going to provide your OpenAI API key so that it can use that 181 00:09:52,890 --> 00:09:54,050 in the deployment. 182 00:09:54,050 --> 00:09:57,250 And you can also give your pushover tokens as well. 183 00:09:57,530 --> 00:10:01,330 And then I have got the rest of the instructions right there for you. 184 00:10:01,810 --> 00:10:06,810 And once that's done, it will be deployed into a hugging face space. 185 00:10:06,810 --> 00:10:08,850 And you can choose whether that is public or private. 186 00:10:08,850 --> 00:10:12,810 You can have it be private if you don't want other people using your key. 187 00:10:12,850 --> 00:10:16,170 Although as I say, it's super cheap, but people would have to. 188 00:10:16,210 --> 00:10:20,690 It would have to go viral for you to have any kind of significant spend. 189 00:10:20,690 --> 00:10:25,450 And of course, make sure that your OpenAI account, you never want to have that auto refilling so that 190 00:10:25,450 --> 00:10:30,210 if that ever were to happen, if your app did go viral, then you would be protected. 191 00:10:30,210 --> 00:10:33,490 And anyway, OpenAI has limits, as you probably know. 192 00:10:34,170 --> 00:10:36,890 So, uh, let's go and actually do this. 193 00:10:36,890 --> 00:10:38,570 Let's run Gradio deploy. 194 00:10:38,770 --> 00:10:43,970 So I'm going to bring up the terminal now, and then we will go and do this ourselves together.