1 00:00:00,000 --> 00:00:01,000 Hello guys! 2 00:00:01,000 --> 00:00:07,000 So welcome to this amazing course on complete generative AI with Lang Chen Ecosystem. 3 00:00:07,000 --> 00:00:13,000 And here our main aim is to specifically focus on building amazing generative AI applications with the 4 00:00:13,000 --> 00:00:14,000 help of Lang Chen. 5 00:00:15,000 --> 00:00:18,000 Uh, here we are just not going to create simple AI applications here. 6 00:00:18,000 --> 00:00:21,000 We are going to probably create Rag applications. 7 00:00:21,000 --> 00:00:23,000 We are going to create search engines. 8 00:00:23,000 --> 00:00:25,000 We are going to create text summarization application. 9 00:00:25,000 --> 00:00:28,000 We are going to create agent type applications and all. 10 00:00:28,000 --> 00:00:33,000 And there are lot many examples that we are going to see, uh, as we go ahead in this entire curriculum. 11 00:00:33,000 --> 00:00:41,000 Uh, I have probably uploaded more than 13 plus projects, end to end projects that will be available 12 00:00:41,000 --> 00:00:42,000 over here. 13 00:00:42,000 --> 00:00:49,000 So end to end projects right now in this video I really want to give you the flow, like how we are 14 00:00:49,000 --> 00:00:51,000 going to learn this entire course. 15 00:00:51,000 --> 00:00:51,000 Okay. 16 00:00:51,000 --> 00:00:55,000 So first of all we will be starting with one programming language. 17 00:00:55,000 --> 00:00:58,000 This is called as Python programming language. 18 00:00:58,000 --> 00:00:59,000 Okay. 19 00:00:59,000 --> 00:01:05,000 Now in this Python programming language, everything that is specifically required in order for you 20 00:01:05,000 --> 00:01:10,000 to get started with this particular course, everything I've actually covered in this Python programming 21 00:01:10,000 --> 00:01:10,000 language, right? 22 00:01:10,000 --> 00:01:16,000 And obviously when I say Python programming language, this is just not like basic type of level. 23 00:01:16,000 --> 00:01:20,000 It is completely from basic to advanced, right? 24 00:01:20,000 --> 00:01:25,000 So here I have actually covered amazing libraries like numpy, pandas. 25 00:01:25,000 --> 00:01:26,000 I've covered inheritance. 26 00:01:26,000 --> 00:01:30,000 I've covered multiple things with respect to this so that you will be able to even file operation is 27 00:01:30,000 --> 00:01:31,000 also covered in this. 28 00:01:31,000 --> 00:01:31,000 Right. 29 00:01:31,000 --> 00:01:35,000 So you'll be able to probably implement this all this kind of projects as we go ahead. 30 00:01:35,000 --> 00:01:41,000 Right then, uh, coming to the second and the, uh, important thing that we are going to learn over 31 00:01:41,000 --> 00:01:45,000 here is nothing but NLP in deep learning, right? 32 00:01:45,000 --> 00:01:51,000 So here we are going to probably also cover this entire NLP with deep learning, because this is something 33 00:01:51,000 --> 00:01:56,000 we really need to know before we go ahead and start working with LM models and multi models. 34 00:01:56,000 --> 00:01:56,000 Right. 35 00:01:56,000 --> 00:02:00,000 So in NLP in deep learning we are going to learn about RNN. 36 00:02:00,000 --> 00:02:02,000 We are going to learn about LSTM, RNN. 37 00:02:03,000 --> 00:02:08,000 Then after covering LSTM RNN you will also be seeing we will be covering GRU RNN. 38 00:02:09,000 --> 00:02:14,000 So in every RNN that we will be seeing, we will understand what is the theoretical intuition, what 39 00:02:14,000 --> 00:02:15,000 is the practical intuition? 40 00:02:15,000 --> 00:02:20,000 What are the problems with the previous one and why the next version of the RNN are specifically coming 41 00:02:20,000 --> 00:02:20,000 up? 42 00:02:20,000 --> 00:02:27,000 And if I probably consider this entire RNN, this is important because this will be useful for solving 43 00:02:27,000 --> 00:02:28,000 any kind of NLP use cases. 44 00:02:28,000 --> 00:02:34,000 Okay, so once we cover the RNN then we will be working with bidirectional RNN. 45 00:02:34,000 --> 00:02:36,000 This is also another variant okay. 46 00:02:37,000 --> 00:02:43,000 Then after getting through all these things then we will be also understanding about encoder decoder. 47 00:02:43,000 --> 00:02:44,000 Right. 48 00:02:44,000 --> 00:02:47,000 And this will have the entire mathematical intuition. 49 00:02:47,000 --> 00:02:51,000 And this will be super important in the interviews because in interviews also they specifically ask 50 00:02:51,000 --> 00:02:52,000 this kind of questions. 51 00:02:52,000 --> 00:02:55,000 And right now also in interviews they focus on basic things. 52 00:02:55,000 --> 00:02:56,000 Right. 53 00:02:56,000 --> 00:02:59,000 So we will be covering with respect to encoder decoder. 54 00:02:59,000 --> 00:03:04,000 Then along with this we will also be seeing something called as attention mechanism. 55 00:03:04,000 --> 00:03:10,000 Now after seeing this attention mechanism we will be seeing some more architecture like Transformer 56 00:03:10,000 --> 00:03:12,000 and Bert. 57 00:03:12,000 --> 00:03:13,000 Right. 58 00:03:13,000 --> 00:03:18,000 So we will also be seeing this now with respect to all these topics. 59 00:03:18,000 --> 00:03:19,000 Once we probably complete it. 60 00:03:19,000 --> 00:03:28,000 right, this will actually give us a basic building block to probably understand about basic building 61 00:03:28,000 --> 00:03:35,000 block, to understand about what it will help us to understand about, uh, LM models. 62 00:03:35,000 --> 00:03:35,000 Right. 63 00:03:37,000 --> 00:03:41,000 LM models, or in short, generative AI. 64 00:03:41,000 --> 00:03:42,000 Right. 65 00:03:42,000 --> 00:03:47,000 So then we will probably go ahead and cover what exactly is generative AI, what exactly we'll be able 66 00:03:47,000 --> 00:03:47,000 to do with this. 67 00:03:47,000 --> 00:03:48,000 Right. 68 00:03:49,000 --> 00:03:55,000 Uh, after understanding about generative AI, then we are also going to go ahead and see about Lang 69 00:03:55,000 --> 00:03:56,000 chain ecosystem. 70 00:03:57,000 --> 00:04:06,000 Right now, Lang chain is one of the most popular web framework, uh, most popular generative AI framework 71 00:04:07,000 --> 00:04:13,000 because this will actually help you to develop amazing generative AI functionalities. 72 00:04:13,000 --> 00:04:13,000 Okay. 73 00:04:13,000 --> 00:04:16,000 API generative AI applications. 74 00:04:16,000 --> 00:04:19,000 Now the best thing about plankton ecosystem? 75 00:04:19,000 --> 00:04:21,000 It has almost each and every thing right? 76 00:04:21,000 --> 00:04:27,000 From this you can actually go ahead and use your both paid LLM models multi model. 77 00:04:28,000 --> 00:04:35,000 Not only that, in our course we are also going to cover open source LLM LM models, open source LM 78 00:04:35,000 --> 00:04:36,000 models. 79 00:04:37,000 --> 00:04:41,000 Uh, and this link chain ecosystem provides each and every functionality. 80 00:04:41,000 --> 00:04:48,000 If you definitely want to work with agents, you want to probably go ahead and create applications with 81 00:04:48,000 --> 00:04:54,000 respect to, uh, you know, uh, let's say one example that I'm going to probably take in this, you 82 00:04:54,000 --> 00:04:58,000 know, we will be going and chatting with SQL database. 83 00:04:58,000 --> 00:04:58,000 Right. 84 00:04:58,000 --> 00:05:02,000 So this you'll be seeing here the chat will basically happen with text. 85 00:05:02,000 --> 00:05:08,000 So this kind of applications will be able to generate our main aim of considering the entire long chain 86 00:05:08,000 --> 00:05:15,000 ecosystem is that because it has many, many features we have vector database right. 87 00:05:15,000 --> 00:05:17,000 We have vector database. 88 00:05:17,000 --> 00:05:18,000 We have retrievals. 89 00:05:18,000 --> 00:05:24,000 What exactly is retrievals will try to understand how to probably go ahead and create chains, how to 90 00:05:24,000 --> 00:05:32,000 do the text summarization right, how to work with creating with chat bots will work with chat bots 91 00:05:32,000 --> 00:05:40,000 and this chat bots will be working with chat message history, chat message history. 92 00:05:41,000 --> 00:05:46,000 So there will be a lot of functionalities that we are going to break down the entire long chain ecosystem. 93 00:05:46,000 --> 00:05:50,000 And in this long chain ecosystem, there are some of the very important modules that we are going to 94 00:05:50,000 --> 00:05:52,000 specifically use. 95 00:05:52,000 --> 00:05:58,000 Uh, if I talk about long chain, there are these three important libraries that we are going to use. 96 00:05:58,000 --> 00:06:03,000 One is lang chain lang chain core. 97 00:06:03,000 --> 00:06:07,000 We are going to go or we will be getting to know more about it. 98 00:06:07,000 --> 00:06:10,000 What exactly is Lang chain core and why this community is specifically used? 99 00:06:10,000 --> 00:06:17,000 Then we will be seeing about Lantern Community, right? 100 00:06:18,000 --> 00:06:23,000 Uh, we will also get to know about this, uh, what exactly what kind of functionalities is specifically 101 00:06:23,000 --> 00:06:24,000 in Lantern Community. 102 00:06:24,000 --> 00:06:27,000 Then we will be seeing something called as lantern tools. 103 00:06:28,000 --> 00:06:28,000 Right. 104 00:06:28,000 --> 00:06:33,000 And there are many more important modules that we are going to use in this entire ecosystem. 105 00:06:33,000 --> 00:06:39,000 But at the end of the day, you will be understanding some amazing ways of creating generative AI applications. 106 00:06:39,000 --> 00:06:41,000 And this will be not just simple projects. 107 00:06:41,000 --> 00:06:46,000 It will be very, very extreme and important projects which will actually help you to handle real world 108 00:06:46,000 --> 00:06:48,000 AI scenario, real world use case scenarios. 109 00:06:48,000 --> 00:06:55,000 Okay, then after probably developing multiple application with the help of link chain, then uh, we 110 00:06:55,000 --> 00:07:00,000 will be going ahead and understanding about some of the deployment techniques. 111 00:07:01,000 --> 00:07:07,000 Now, this deployment techniques that I am going to specifically use, I'm going to make sure that whatever 112 00:07:07,000 --> 00:07:12,000 free resources are available, right. 113 00:07:13,000 --> 00:07:16,000 Resources are available with respect to deployment. 114 00:07:16,000 --> 00:07:17,000 I will be doing that. 115 00:07:17,000 --> 00:07:18,000 Okay. 116 00:07:18,000 --> 00:07:22,000 I will not take any paid kind of things with respect to this particular deployment techniques. 117 00:07:22,000 --> 00:07:22,000 Okay. 118 00:07:22,000 --> 00:07:26,000 So all these applications also I will talk about it guys. 119 00:07:26,000 --> 00:07:32,000 We will be using Streamlit to probably develop this entire application. 120 00:07:32,000 --> 00:07:35,000 Streamlit is another framework that we specifically use right. 121 00:07:35,000 --> 00:07:43,000 So here also we are going to specifically use this Streamlit Streamlit library okay. 122 00:07:44,000 --> 00:07:52,000 Now after covering deployment techniques uh, obviously we also need to know about fine tuning techniques. 123 00:07:53,000 --> 00:07:59,000 So in this fine tuning techniques we are going to see some of the tuning things, uh fine tuning techniques 124 00:07:59,000 --> 00:08:00,000 here. 125 00:08:00,000 --> 00:08:04,000 Uh, we will be specifically using Google Colab. 126 00:08:04,000 --> 00:08:04,000 Again. 127 00:08:04,000 --> 00:08:08,000 We will try to just use free resources and try to do this. 128 00:08:08,000 --> 00:08:09,000 Okay. 129 00:08:09,000 --> 00:08:13,000 Uh, one of the paid LM model will be OpenAI. 130 00:08:13,000 --> 00:08:13,000 Okay. 131 00:08:13,000 --> 00:08:15,000 And here I will also talk about what? 132 00:08:15,000 --> 00:08:17,000 Uh, uh, LM models. 133 00:08:17,000 --> 00:08:18,000 We are specifically going to work. 134 00:08:18,000 --> 00:08:19,000 Okay. 135 00:08:19,000 --> 00:08:24,000 So over here we are just going to use Google Colab, and we are going to learn about some of the important 136 00:08:24,000 --> 00:08:26,000 concepts and techniques. 137 00:08:26,000 --> 00:08:29,000 Uh, like quantization. 138 00:08:30,000 --> 00:08:31,000 Quantization. 139 00:08:31,000 --> 00:08:35,000 We are going to learn about Laura Clara. 140 00:08:35,000 --> 00:08:43,000 And with the help of this, we are going to fine tune, fine tune with some open source LM models. 141 00:08:43,000 --> 00:08:46,000 Open source LM models. 142 00:08:46,000 --> 00:08:47,000 Okay. 143 00:08:48,000 --> 00:08:50,000 Open source LM models. 144 00:08:50,000 --> 00:08:56,000 Okay, so we are also going to cover this so that you also get some experience with respect to the fine 145 00:08:56,000 --> 00:08:57,000 tuning things. 146 00:08:57,000 --> 00:08:57,000 Okay. 147 00:08:57,000 --> 00:09:02,000 Now coming to the sixth topic, uh, sixth important module, which we are basically going to go ahead 148 00:09:02,000 --> 00:09:09,000 and create, you know, it is working with Gen I in AWS, right? 149 00:09:09,000 --> 00:09:15,000 So let's say if you clearly want to completely work or develop your generative AI applications in AWS 150 00:09:15,000 --> 00:09:18,000 by using AWS bedrock. 151 00:09:19,000 --> 00:09:25,000 So there is, uh, there is a service in AWS bedrock which will have many, many open source models 152 00:09:25,000 --> 00:09:28,000 that will be available over there in the hosted mode. 153 00:09:28,000 --> 00:09:28,000 Right? 154 00:09:28,000 --> 00:09:35,000 So with the help of this AWS bedrock and using other important services that are present in AWS, like 155 00:09:35,000 --> 00:09:39,000 uh, Lambda function okay. 156 00:09:39,000 --> 00:09:44,000 So lambda function, then you have API gateway. 157 00:09:45,000 --> 00:09:49,000 So here we will be developing some amazing end to end projects. 158 00:09:50,000 --> 00:09:51,000 End to end projects. 159 00:09:51,000 --> 00:09:57,000 And we will also see how we can deploy this entirely in our AWS SageMaker. 160 00:09:58,000 --> 00:09:58,000 Right? 161 00:09:58,000 --> 00:10:00,000 AWS SageMaker. 162 00:10:01,000 --> 00:10:05,000 So this will be one another amazing module that we are going to work with. 163 00:10:05,000 --> 00:10:05,000 Okay. 164 00:10:06,000 --> 00:10:13,000 In the recent days, uh, one more module is becoming very much popular that is called as Nvidia Nim. 165 00:10:14,000 --> 00:10:14,000 Okay. 166 00:10:14,000 --> 00:10:16,000 Nvidia Nim. 167 00:10:16,000 --> 00:10:24,000 Now Nvidia name uh, is just like, uh, you have this this Nvidia name actually helps you to deploy 168 00:10:24,000 --> 00:10:27,000 the generative AI models in the cloud. 169 00:10:27,000 --> 00:10:32,000 And here with the help of Nvidia name, we will be seeing what all kind of open source models are already 170 00:10:32,000 --> 00:10:35,000 available in this, uh, hosted platform itself. 171 00:10:35,000 --> 00:10:40,000 And with the help of this also we will be trying to create some end to end projects. 172 00:10:40,000 --> 00:10:43,000 And here also we are going to use Lang Chain. 173 00:10:44,000 --> 00:10:44,000 Okay. 174 00:10:44,000 --> 00:10:47,000 We are specifically going to use Lang Chain. 175 00:10:47,000 --> 00:10:55,000 So as you know as this is 2024 we definitely need to new all the new techniques that are available in 176 00:10:55,000 --> 00:10:57,000 the market which will be definitely helpful for you. 177 00:10:57,000 --> 00:10:58,000 Okay. 178 00:10:58,000 --> 00:11:06,000 And finally, uh, I'm also going to cover one more amazing library, which is called as Crew Eye Crew. 179 00:11:06,000 --> 00:11:10,000 I actually helps you to create multi eye agents. 180 00:11:11,000 --> 00:11:18,000 Multi agents, which helps you to solve different kind of which helps you to develop different kind 181 00:11:18,000 --> 00:11:20,000 of generative AI applications. 182 00:11:20,000 --> 00:11:25,000 And this also we will try to use long chain and we'll develop this okay. 183 00:11:25,000 --> 00:11:31,000 Along with this particular AI framework okay I think over here Lange chain will not be used but it has 184 00:11:31,000 --> 00:11:32,000 its own framework. 185 00:11:32,000 --> 00:11:36,000 So with the help of this also we will be generating some kind of amazing applications. 186 00:11:37,000 --> 00:11:42,000 Uh, so overall this is the entire course curriculum which I have actually written step by step how 187 00:11:42,000 --> 00:11:43,000 we are going to cover. 188 00:11:43,000 --> 00:11:49,000 But I feel after doing this many things, uh, you should definitely be able to crack any kind of jobs 189 00:11:49,000 --> 00:11:51,000 and even open your own startup. 190 00:11:51,000 --> 00:11:57,000 Specifically working with any kind of models that you really want to work with right here. 191 00:11:57,000 --> 00:12:04,000 If I just consider the LM models that I am actually going to use, as I said, LM and multi model. 192 00:12:04,000 --> 00:12:10,000 Okay, as I said, I'm going to use both paid and open source so that you get an idea with respect to 193 00:12:10,000 --> 00:12:12,000 both paid and open source. 194 00:12:12,000 --> 00:12:15,000 The first one I'm going to specifically use open AI. 195 00:12:15,000 --> 00:12:15,000 OpenAI. 196 00:12:16,000 --> 00:12:20,000 In OpenAI, I will be showing you examples with uh, with like GPT four. 197 00:12:20,000 --> 00:12:21,000 All right. 198 00:12:21,000 --> 00:12:28,000 We will be seeing with respect to GPT four turbo, we'll be seeing some OpenAI embedding techniques, 199 00:12:29,000 --> 00:12:32,000 OpenAI embedding techniques. 200 00:12:32,000 --> 00:12:33,000 Right. 201 00:12:33,000 --> 00:12:39,000 When I go to open source model, I will be specifically using this Google Gamma model. 202 00:12:39,000 --> 00:12:42,000 So I will be using this Google Gamma model. 203 00:12:42,000 --> 00:12:44,000 Gamma two model. 204 00:12:44,000 --> 00:12:45,000 Right. 205 00:12:45,000 --> 00:12:51,000 Along with this, you'll be seeing the third important open source model that I will be using will be 206 00:12:51,000 --> 00:12:51,000 from meta. 207 00:12:52,000 --> 00:12:56,000 So obviously models like llama three I'll be using it from here. 208 00:12:56,000 --> 00:12:59,000 Along with that and Mistral. 209 00:12:59,000 --> 00:13:03,000 So if you know about the company called as anthropic right. 210 00:13:03,000 --> 00:13:05,000 We're we're also going to use this particular model. 211 00:13:05,000 --> 00:13:10,000 That is nothing but Mistral from anthropic, which is again an open source model that is available. 212 00:13:10,000 --> 00:13:15,000 One of the version of this Mistral, uh, we will be using to solve some of the problems. 213 00:13:15,000 --> 00:13:15,000 Right. 214 00:13:15,000 --> 00:13:22,000 And uh, along with this, uh, uh, you know, in Meta Llama three and we are also going to use something 215 00:13:22,000 --> 00:13:23,000 called as code llama. 216 00:13:23,000 --> 00:13:24,000 Okay. 217 00:13:24,000 --> 00:13:27,000 Code llama is specifically used for creating code assistant. 218 00:13:27,000 --> 00:13:27,000 And. 219 00:13:27,000 --> 00:13:28,000 All right. 220 00:13:28,000 --> 00:13:32,000 So GPT four or GPT four turbo gamma model this model. 221 00:13:32,000 --> 00:13:34,000 And then finally obviously we are also going to use hugging face. 222 00:13:34,000 --> 00:13:36,000 How can I miss this. 223 00:13:36,000 --> 00:13:37,000 Right. 224 00:13:37,000 --> 00:13:44,000 So there are a lot of models uh LM models that are specifically available both open source and paid 225 00:13:44,000 --> 00:13:45,000 in the hugging face. 226 00:13:45,000 --> 00:13:47,000 So we are going to specifically use this. 227 00:13:47,000 --> 00:13:52,000 And at the end of the day, by using all this kind of models, we are going to probably create a lot 228 00:13:52,000 --> 00:13:53,000 of generative AI applications. 229 00:13:53,000 --> 00:14:00,000 I will also introduce you how you can actually use this entire open source model with the help of grok 230 00:14:00,000 --> 00:14:04,000 AI infrastructure, which is important for everyone. 231 00:14:04,000 --> 00:14:06,000 Grok infrastructure. 232 00:14:06,000 --> 00:14:06,000 Right. 233 00:14:07,000 --> 00:14:09,000 Um, grok is an LPO engine. 234 00:14:09,000 --> 00:14:10,000 Okay. 235 00:14:10,000 --> 00:14:10,000 We will talk about it. 236 00:14:10,000 --> 00:14:14,000 What exactly is this LPO engine and how it is better than the GPUs? 237 00:14:14,000 --> 00:14:15,000 Okay. 238 00:14:15,000 --> 00:14:19,000 So overall this is the entire idea of the entire curriculum itself. 239 00:14:19,000 --> 00:14:24,000 There are many things that needs to be discussed, and you will be seeing that as you go ahead. 240 00:14:24,000 --> 00:14:28,000 And all these sessions are like completely in depth, explained. 241 00:14:28,000 --> 00:14:33,000 Um, every project is developed in such a way that you should be able to implement it. 242 00:14:33,000 --> 00:14:35,000 I've written each and every line of code as we go ahead. 243 00:14:35,000 --> 00:14:36,000 Right. 244 00:14:36,000 --> 00:14:38,000 So I hope, uh, you like this particular video. 245 00:14:38,000 --> 00:14:40,000 I will see you all in the next video. 246 00:14:40,000 --> 00:14:40,000 Thank you. 247 00:14:40,000 --> 00:14:41,000 Take care. 248 00:14:41,000 --> 00:14:41,000 Bye bye.