1 00:00:00,400 --> 00:00:04,320 Okay, now it's on to a proper example that we'll use with Landgraaf. 2 00:00:04,680 --> 00:00:05,440 Just take the obvious. 3 00:00:05,440 --> 00:00:10,040 The reason I showed you the silly example was because I wanted to show that nodes don't need to have 4 00:00:10,040 --> 00:00:13,320 calls to llms, and they still do what they're meant to do. 5 00:00:13,360 --> 00:00:15,360 But now we are going to add an LLM. 6 00:00:15,400 --> 00:00:17,880 So we start by defining the state. 7 00:00:18,080 --> 00:00:21,040 We create a graph builder with that state. 8 00:00:21,160 --> 00:00:26,480 And now we create an LLM a real LLM using chat OpenAI. 9 00:00:26,520 --> 00:00:31,840 So chat OpenAI is a construct from Lang Chain the sibling to Landgraf. 10 00:00:32,000 --> 00:00:35,520 Uh and that's what we'll be using to connect with our LLM. 11 00:00:35,680 --> 00:00:39,640 And now you don't need to use lang chains llms for this. 12 00:00:39,640 --> 00:00:41,120 You can use any Llms. 13 00:00:41,120 --> 00:00:43,160 You could directly call the LLM yourself. 14 00:00:43,320 --> 00:00:47,560 You could also, uh, use maybe OpenAI agents SDK. 15 00:00:47,840 --> 00:00:53,320 But it does make things a bit simpler sometimes if you use Lang chain and most of the community examples 16 00:00:53,320 --> 00:00:55,840 of course go from Landgraf to Lang chain. 17 00:00:55,840 --> 00:00:57,520 So it's it's easy to do it that way. 18 00:00:57,520 --> 00:00:58,880 And that's what we'll do for here. 19 00:00:59,440 --> 00:01:03,150 So we're going to create a new node called chatbot node. 20 00:01:03,150 --> 00:01:06,070 It takes an old state and it returns a new state. 21 00:01:06,230 --> 00:01:07,390 And what does it do. 22 00:01:07,430 --> 00:01:12,390 Well it takes the the LM and it invokes on that LM. 23 00:01:12,390 --> 00:01:13,830 So it's again it's the Lang chain. 24 00:01:13,830 --> 00:01:18,950 Landgraf word invoke uh, passing in the messages from old state. 25 00:01:18,950 --> 00:01:21,950 So old state has a messages field and that is what we pass in. 26 00:01:22,270 --> 00:01:29,710 And then for the new state, it creates a new state object which contains within it as in its messages 27 00:01:29,710 --> 00:01:31,790 field, it contains the response. 28 00:01:31,790 --> 00:01:34,230 And we return the new state. 29 00:01:34,510 --> 00:01:40,310 And we add that node called chatbot, uh, into our graph builder. 30 00:01:40,350 --> 00:01:40,830 Done. 31 00:01:41,190 --> 00:01:47,150 Now we'll add some edges from start to chatbot from chatbot to end done. 32 00:01:47,510 --> 00:01:49,710 And now we will compile our graph. 33 00:01:49,710 --> 00:01:51,510 Step five and we'll look at the graph. 34 00:01:51,510 --> 00:01:54,750 And sure enough it goes start to chatbot to end. 35 00:01:55,350 --> 00:01:59,070 And then we put it all together in a simple gradio chat function. 36 00:01:59,070 --> 00:02:05,260 It takes an initial state, which is a state object set up with these messages like so. 37 00:02:05,700 --> 00:02:09,620 We then call graph dot invoke to actually call our graph. 38 00:02:09,740 --> 00:02:16,260 We print the result and we will also show the results back in Gradio. 39 00:02:16,540 --> 00:02:17,460 So here it is. 40 00:02:17,500 --> 00:02:18,860 And I can say hi there. 41 00:02:20,300 --> 00:02:22,180 And it's actually calling OpenAI. 42 00:02:22,220 --> 00:02:25,260 Now it's not using our silly adjectives. 43 00:02:25,420 --> 00:02:32,540 And you'll see down here that there's the user message and the response coming in these objects, this 44 00:02:32,540 --> 00:02:35,460 human message object that is coming back. 45 00:02:35,460 --> 00:02:36,780 Sorry, that human message is my. 46 00:02:36,780 --> 00:02:37,340 Hi there. 47 00:02:37,380 --> 00:02:38,980 I mean, there should be an AI message. 48 00:02:38,980 --> 00:02:39,740 There it is. 49 00:02:39,940 --> 00:02:44,740 There is the AI message, which is the response actually coming from OpenAI. 50 00:02:45,420 --> 00:02:50,860 But one thing that's worth noting is that if I continue this conversation every time we are invoking 51 00:02:50,980 --> 00:02:56,660 this graph, and you will see what you have probably already suspected, which is that we're not actually 52 00:02:56,700 --> 00:02:58,740 keeping track of any history here. 53 00:02:58,970 --> 00:03:00,410 Let's see that in action. 54 00:03:00,610 --> 00:03:02,770 Uh, if I, uh, say, um. 55 00:03:03,010 --> 00:03:03,890 My name's Ed. 56 00:03:07,490 --> 00:03:08,410 Nice to meet you, Ed. 57 00:03:08,450 --> 00:03:09,770 How can I assist you today? 58 00:03:11,290 --> 00:03:12,290 What's my name? 59 00:03:15,050 --> 00:03:17,770 I'm sorry, but I don't have access to your personal data. 60 00:03:17,970 --> 00:03:21,450 So there's a sign that it's not able to keep context. 61 00:03:21,450 --> 00:03:26,370 And you can see it yourself if you read the information that's going to and fro. 62 00:03:26,530 --> 00:03:33,050 So because we've just got this simple graph that we were invoking each time, there's nothing particularly 63 00:03:33,050 --> 00:03:34,770 interesting happening here. 64 00:03:34,810 --> 00:03:41,370 And the state uh, is just, just contains that that uh, doesn't contain the history or anything. 65 00:03:41,610 --> 00:03:44,850 So, uh, that's one of the things that we clearly need to address. 66 00:03:45,170 --> 00:03:47,370 And the good news is that we will indeed address it. 67 00:03:47,370 --> 00:03:49,450 But the bad news is not until tomorrow. 68 00:03:49,570 --> 00:03:56,170 But we'll also address things like tools, our old favorite, along with a couple of other things. 69 00:03:56,330 --> 00:03:58,010 So, uh, look forward to it. 70 00:03:58,010 --> 00:03:58,810 I'll see you then.