GenAI Course

Transcription for the video titled "GenAI Course".


Note: This transcription is split and grouped by topics and subtopics. You can navigate through the Table of Contents on the left. It's interactive. All paragraphs are timed to the original video. Click on the time (e.g., 01:53) to jump to the specific portion of the video.


Intro (00:00)

We generally talk about some a topic and try to look at some code, try to make some applications or something. So basically I wanted to conduct this session to actually see to give a detailed plan for anyone who is just starting out building with AI. Right. So here is how the session will go. So yeah, this is a scheduled one hour session, but I'll probably take only 20 to 25 minutes initially.

Exploring Ai: Projects, Concepts, And Takeaways

My First Journey in AI (00:28)

To talk about what are all the things that you can do when starting out with A and probably the next 25 minutes or so or the remaining remainder period of time. The floor will be open for any questions, right? So if you have any doubts regarding whatever is presented during the session, or if you know wish to get some of your specific questions answered, this is probably the best time to do it. So as always for anyone who is attending these sessions for the first time, so I'll just give a brief introduction about myself. So yeah, I'm an IT, Delhi 19 graduate, so around for 3 1/2 to four years. I've worked as a data scientist in various multinational banks where I was very much into risk analytics and other sort of stuff. So I have a particular interest in the education space. So that is when I founded an NGO back when I was in college at IIT. And currently, yeah, continuing on my interest in education as well as a, so I'm currently founded the build faster there. So this is a platform that will help you build with air, right. So it's a collaborative community for builders. So we will have some workshops, courses in many other resources as a part of these things. So we have conducted close to 10 plus workshops so far. So cumulatively we have seen around more than thousand people have attended the session so far till now. So with respect to the courses and all, we'll soon be launching some full length courses. So you can stay tuned like stay tuned for this. So I think I just forgot to say, so there is a WhatsApp group that we have, so the link is in the chat if you have not, if you are not the member of the group already. So you can just click on that link and join WhatsApp group for any updates. So all the updates about the workshops, courses and the other resources that I generally keep share, so I'll just do it on WhatsApp and the Mailer list that I have, right? So if you're not doing the WhatsApp group already, I think you can do it right now. So what about next event? So we will have a small session next weekend as well, but I wanted to just announce the master class that I'll be conducting on 26th August, which is like 2 weeks from now. So today is 12th August, right? So we will have the session on 26th, which is the next Saturday. So I'll just give a brief of what will happen in the session and I'll give you more details as to how to register at the end. But it'll be a three hour workshop on how to build chat bot on your data, right? So you can upload your PDF file, you can upload your text file or you can even upload your video file. And then start chatting with them. So I'll also show you a demo of what you will what you can expect from this particular session. So stay tuned for it. So like stay till the end. Do actually get more details on how to register for this workshop. So having said that, so let us just proceed with the session today. So I think there was some chat. So are there any questions? OK, great. So now let us just start with today's presentation which is beginners road map to AI. So whenever starting to make this particular presentation, so my general thought process was so let me just think of. What are all the steps Step 123 so on for? Let me just create a road map and then try to present it, right So here. We have people from various backgrounds, right? Someone is a developer with five years of experience. Someone is a product manager with two years of experience. Someone is an experienced professional with more than 15 years of experience. So we have all sort of people in the group for here. So giving a specific step by step guide would not be a very, you know, a great thing to do and would not like that would be a very generalized thing to do. So instead what I thought I will do is I'll just show you. My personal journey as to how I started with A and how I started developing these applications. So maybe I'll just spend 10 minutes to give an idea of how I started my journey into AA. So what are the resources that I generally used and what are the questions that I had initially and how did I get answers for them? Right. So at every step I'll just tell you what my logic was in terms of how to go about the journey. And maybe those my journey will probably give you some insights as to how you can think about your journey per SE. And I will also introduce you to some tools and resources where you can, you know, get latest updates on AI and maybe some tools to help you write code better or maybe so on and so forth, right. So we'll spend around 10 to 15 minutes on the initial My Journey part and then the next 10-15 minutes on the tools and resources part. Great. So I think I have gone through these four stages right? So while thinking about my AI journey. I think I have gone through these four stages starting from ignoramus to follower to explorer to builder. Right So I just wanted to give a brief understanding of what this is.

Ignoramus, Follower, Explorer, Builder (05:56)

So ignorance is obviously like. I didn't want it to use ignorant word so because that is probably a sensitive word to use but otherwise so initially I was so you might think that I started in IIT and I did maths and computing so I was already well versed with a so that is far from true. So I had basic idea of I had very basic understanding of. ML models, so not anything more than that. And I was not even a follower or something, right. So I was not more or less uninterested in this particular domain at least till few months back, let's say maybe even till like let's say one year back or something. So this is probably where I was one year or maybe eight months back. This is what I was. So when ChatGPT got released or general APS got released, that was when I started following the sector very closely. Right, so instead of instead of just. Sitting idle, I was just trying to explore all the other tools in terms of what ChatGPT is and I was just following. People here and there to get some news and I was just following and nothing. So there was nothing that I was actively doing to pursue air, right. So I was not doing anything actively to pursue air. So then when I thought that A is going to be a very big thing and this is probably an inflection point in how people deal with technology, that was when I quit my job a few months back, around 3 months back. And that was when I got into the explorer zone of the things right. So instead of just passively observing the updates or in this sector, I was actively following it as to what are the updates that are happening on a daily basis or a weekly basis. So what can I learn, how can I learn? So on and so forth, right. So you can just think that I spent in this particular zone for around one to two months. So since I had no job to do, and since my main intention was to build with a, so I started exploring what are the things that I can do to become a builder, right? So by builder, again, as always, I just don't mean the developer side of things, right? So builder means if you're wanting to conceive, any ideas, right? If you're wanting to conceive any AI ideas, you need to have some sort of a framework in your mind, right? So this was the face like let's say during the explorer phase.

Putideas (08:23)

Actively exploring things as to water all the things that are happening, so on and so forth. Yeah, and following a lot of different things and then slowly understanding how can I build these tools myself. So this was my journey so far, right? And if you are in this particular session, so you might be in any of these three sessions or any like you might be in any of these four sections, right? So the suggestions that I'll give you will in this particular session will help you move from one stage to another, right? So if you are someone who is just starting out with no idea of what. Anything is happening, then fine. So if you are someone who is just passively following the sector thinking to jump into this sector but don't know where to start, yeah you will still get some suggestions for this. And yeah, like even if you are an explorer or the builder, you will still be able to, you know, get some resources and tools that you can use later on. Right, so now. So now. I think this is a basic journey that I went through in terms of. Starting without nothing and making apps so I'll tell you like what my context was right so how my journey was. So just to give a context as to So what was my initial level of understanding about air coding initially. So before starting my journey right so before going through this sorry. So before going through this journey, I'll just tell you, set up the context so that these were my attributes before starting. So what you guys can do is you can think of what are your, what is your context before starting, right? So you can think about what are you good at, what is it that you want to learn, so on and so forth, right. So while I was starting so I had a very basic understanding of some ML models in terms of some regression models, some sort of random forest models, very simple models, nothing too complicated. I had a very basic understanding of these models. So I was fairly good at Python so I was making a lot of data dashboards or something similar to it as a part of my previous job. So I was more or less comfortable with coding and I had an experience of working with large amounts of data before so this was where I was at. So in my journey this was what my initial context was. So starting with then with these things in my mind, where did I start? Tell me where do I start? So what I did was I watched some countless YouTube videos. Let me get into the theory part of it. So I started reading some blogs, I started reading some papers, so on and so forth. More often than not. More often than not, I was not able to understand more than, let's say 30 to 40% of what is being written in the paper. Because those papers or blocks were extremely theoretical. And even if let's say I have understood 30 or even 50% of the paper, so there was no where to go with that, right. So I had an understanding of. What back propagation is? But what do I do with that? Right, So there is nothing that I can do. And one other thing was that I spoke to my friends from IIT or otherwise who are working in LNP and this was one of the most important thing because that was the time when I realised no one knew anything. Right. I spoke to one of my friend who had around 3 plus years of research experience in MLP. So when I asked him. How do I get started with GPT? How do I get started with LLMS, so on and so forth? He didn't have an answer, right? The reason was because all the technology that you are seeing right with respect to GPTLLMS is only a few months old. Right. So GPT was introduced let's say in December or something and the APIs got released around the five to six months back. So even if there is someone who is researching in this domain for a lot of years, they didn't have a clear Rd. as to how to go about this thing, right? So that thing was very clear to me then you will have a very significant beginner advantage if you start with AI today. You will get a very significant advantage. It's because the people who are working in this domain for 10 years or 15 years at a time, of course they will have a better understanding of how the architecture is, how to build better tools, so on and so forth. But all the tools that are coming around, yeah, or all the models that are coming around are only a few months old, right. So you are not going to like get any significant advantage from the people who are starting now. And this was also one of the main reasons has to take a full time jump into this thing. Because all the things about a are pretty new and if you wanted to start now, this is the best time because you will get a significant advantage here, right? So this is where I was, where I was trying to explore a lot of things here and there.

Prompt Templates at Langton Tutorials (13:53)

But I'll tell you what are the resources that are used to start in, let's say, the first month, right? So maybe a few months back, three to four months back. This was the first, you know, most useful tool that I've encountered, right? So these were the Langton tutorials put forth by Pine Cone right? So they had around 7 to 8 different lectures. On how to use AI in different contexts. Right, How to use. How to use GPT in different contexts? So these were the first tutorials which actually gave me confidence in terms of. Thinking about an idea and going and actually building it, right? So maybe I'll just maybe open it and show you guys. So let's say if I go to the examples here so I'll send this link to you guys once the session is over. So this is the GitHub link for it and there are around 10 tutorials here that you can see. So these cover some basic parts of what line chain is, how to use different models, what are prompt templates. So maybe I'll just write down here, right? So how to use, let's say? How to use APIs? Some basic understanding of what prompts are, what chains are, how to use different models. You know how to different agents, so on and so forth. Basic understanding of that thing. I would attribute that to this particular tutorial list, right? And there is also a YouTube channel called by James Bricks I guess. So this was the guy who created all these tutorials and it put forth here, so I'll just show it to you again. Right? So this is the basic intro template. Everything is a simple notebook, right? So what you can basically do is just go click on any tutorial that you want. You don't have to do anything. You don't have to install anything. You don't have to do anything. You just need to have a Gmail account. So once you open just click on the open and collab part. Open in the collar part and you have everything ready, so you don't even have to write anything. Just go and click. Slow and run this particular book and keep reading everything. Like keep reading everything, try to understand each and every step. So one suggestion to do is let's say if you are starting to use these things, so never run the books by themselves, right? So what you can do is open a new notebook, right? Don't run their own notebooks, just open this in the other window. Just take a new collab notebook, right?

Old note (16:45)

Just take a new collab notebook. Maybe just. Maybe you just say new notebook? So you just need a Gmail account for it. You don't have to go through any installation or something. Just open this notebook in the other screen. Or maybe just open it in the other window and try to write everything by. Basic idea for this, right? So the first most important tool was to understand language and basically so that will give you a significant confidence boost in terms of how we can start.

Python learn online (17:37)

So I think there are a lot of people in the group who are interested in no code tools, right? So you might not be interested in like you might not know Python at all. Right. You might not know Python at all. You might not know any coding language per Southeast. If you're interested in no code, you can still go ahead and start doing it right, so maybe I'll just give you some resources for this, but. Just to give an idea, Python learning is not very hard even if you let's say. Have not coded any time in your life. This is still very easy to follow, right? Just open this book, open this particular thing. Even if you're not understanding. Even if you're understanding, let's say 40 to 50% of things for the very first time, that is more than enough. Right. So you can do it multiple times. And if you just go ahead and read these particular notebooks, you can see everything is written in plain English, right? So there are clear structure written to it. There is a very clear structure that is given and. You can basically understand this, not difficult to understand. So first step you can start or maybe I started here, maybe I would suggest you guys to do this as well. Just start with line chain, try to get basic understanding of Python And land chain and then go about it. And one other thing was while I was trying to do all these things, I didn't have a clear understanding of how these models are working. So that is where you know Anders Karpathy is the Tesla's computer vision head, right? And he's also very reknown scientist. He has his own YouTube channel. Yeah, if I'm right, yeah, this is the channel, right? And you just go and watch his videos on language modeling. It will give you a very clear idea as to how he's conveying thoughts, right? Just a basic understanding of how language modelling works, how GPT works, so on and so forth. It will give you a very clear idea. So I started here then once I got a basic confidence, what I did was to develop my first step right. So I'll show you what app I've developed. So this was the app that I've developed where you can just upload any YouTube video and you can just start asking questions to it.

Everyone (19:56)

So this was the first a application that I've ever developed. This I have developed there on 2 and half three months back I guess, right. So there are some version problems. So it might not work right now, but I can, you know modify this and send it to. So this was the first AAF that I've developed. So the basic understanding that I got from these particular. So these were the two main things. But otherwise I have referred to many other sources as well. So what I would suggest is if you're starting to refer these, great but they don't just stick to a single resource, right? So if you don't understanding something which is written in this thing, just go to Google or maybe go to ChatGPT and ask. I don't understand this particular piece of code. Can you explain? Right. So whenever you're looking at these particular resources, just go ahead. Try to replicate the results by yourself and try to understand each and every step as much as possible. And one other important thing is don't force yourself to understand 100% every time, so that is not actually possible right? Even if you don't understand. Let's say 30 to 40% of it. It's OK. Don't force yourself to understand everything in a single go, which will significantly slow your speed and motivation. So I think I have made this first YouTube app. So what I've learned this is that so in the basic first part. So I was already like decently good at Python. So no problems with Python initially, but with plan chain and other AI tools, AI libraries. So this was the understanding that I found initially, right? So now with this Python And land chain as a basic understanding then I like I started to build some UI using streamlet. Streamlit is again a fully Python library. You don't need to write any HTML, CSS, JavaScript. You don't need to learn anything. You don't need any of these things, so Streamlit is an UI library that will help you host your Python apps for free. So here I've made this particular app. Everything is made in Python, so no other language right? So in this particular main learning was now that I had basic understanding of Python And basic AI libraries. So that was when I started building something. Building an app which I could share with other people. Right. And I think there was a question.

Opino's (22:37)

Patrick, hi Yuvraj is right. So as you're saying like learning Python will be very important for us to take that next step and is it? So if again so if you are, if your main goal is to develop something by yourself, then Python is an indispensable tool. So even if JavaScript it will work OK, no being a non coder, anything will be learning from zero. So what would you suggest in that case? Like getting started on Python is better or anything else that I have to focus on. So if you are, let's say if your motive is to build some applications. No code also has a significant learning curve, right? True, it's just that, it's just that it does not involve code. But personal opinion is that I found no code to be difficult than Python, my personal thing. So everything is significant learning curve, but if you are looking to start developing something, then go ahead with Python. Python is as simple as learning English language. So, and just to inform, I'll be launching a Python course so very soon, maybe in the next one month.

Building more apps (23:50)

Not one month, maybe before that. I'll be launching A1 hour course on it to help with this thing, so you can watch out for that. Great. Thanks. Thanks. So basically we have what you're saying is also true because learning something like bubble and all, you can't actually develop till you understand how to connect with APIs. You need to have some basic knowledge. So it's better to learn something from scratch. Yeah. Makes sense. Yes. Thanks. Yes. OK, so I think once I developed this first application, then what next? So now I had a basic understanding of general tools. Then let's say I knew how to build the basic apps, then the next step is to game up, right? So I have developed many different applications, so on and so forth. Around 20 to 25 applications in different domains, in different domains using key, let's say. I'll show you one app, right? So this is one app that I've developed where I have. Created a on different characters right? So here you can just chat with Elon Musk if you wish. So you can ask him are. In this ask him are you friends with? Jeff Bezos or something. So here I'm chatting with Elon Musk. We generally know that Elon Musk and don't see eye to eyes. So here is what he's replying, right. So we have sleepovers, play video games, just typical billionaire romance. So just some witty response, so you can go ahead and talk to SO critics about this thing. So this was the 2nd application that I've developed. So my first application was little bit of basic. Then I stepped up in terms of you know, developing some conversational chains. Then I tried to develop some agents which will help, you know, do data analysis by themselves. And there are some other vector database applications that I've developed, so on and so forth, right. So first step, once you have a basic understanding, once you're confident that you will build basic applications, then go ahead and. Try to have some wild ideas and try to build them and again, it's going to be it's not.

First hands experience riel t"[6]" sliders 1853,Ai transforms interview questions Audrey" (25:55)

It'll not be a very easy journey in the beginning, right? So in the first few weeks of starting, you might not even understand where you are going. But from my personal experience, I'm telling you this is not very hard thing to do, right. So yeah, I think I'll share some main insights from my learning before. Jumping into some generic advice, right. So the main insights is that and which is almost indispensable is that my most learning happened while I was building something, so it didn't happen. Let's say it didn't happen while I was doing launching tutorials. It didn't happen when I was watching Android Karpati's videos. It didn't happen when I was reading some amazing blog posts or something. So most of my learning happened when I was building some apps like this. Right, so that was the main insight from my journey, right? And one thing that you need to do is develop many apps, mini apps or even micro apps let's say, right? And the reason I even in these particular sessions I generally. Stick myself to very basic apps like e-mail generator or tweet generator or something. So these were the apps that I generally covered in previous sessions because once you have confidence that you can build basic stuff. Then you can step up, right? So if you're just starting out, try to develop some mini apps. Let's say something like a simple To Do List. Something like an e-mail generator, Something which involves a basic API operation. Nothing more than that. And absolutely don't jump into fine tuning or open source models or something. So this is the mistake that I've seen a lot of people who are starting to do, right? So on day one itself, they Start learning how to fine tune a LAMA model for their application or something, right? How do I use open source model to protect my data for something else, something else, something else, right. So don't do this like if you are someone who you are just starting out. Who does not understand the basic elements of how things work? Don't go into all these things right now. So even before jumping into fine tuning or something, I just took around two months before I even started exploring these particular parts, right? And I still don't know most of how like how open source models work or something like this, right? So if you are someone who is starting out, don't go anywhere near the fine tuning models, open source models, and these are very fancy to look right? So these might be very fancy to say that I know how to fine tune or something. But on a long run, if you're not able to understand the basics or create basic things then it is not going to take you far. So what is the summary again? Right, So let me just give a quick summary. Maybe you all signed up for a road map. So again there is no step by step road map, but this is what you can do, right? So first thing that I would suggest everyone to do is write down the intent right? Why are you learning this? So my intent was that I want to work full time with AI, want to develop some AI applications in education. And I need to start building it. This was my intent. So writing intent is I think important because it will help you stay motivated throughout the journey, right? If you don't have an intent, if you just start by exploring some random things, once you encounter something which is hard, or once you encounter something which is difficult, you just give up, right? So I think first important step to do is to have a clear intent as to why you want to do right. It might be that you are working on a product. Within your company to help something else or you have XYZ hobby project that you have wanted to work on for a while now and that is where you want to start. So you can write down your intent as to what is it or why is it that you want to learn this particular thing. And next thing is to start small. So this is what I've been telling all the time, right? So start small. Just learn. Just learn Python. Just learn launching basics. Something like that. Just start with this. Then go ahead, use Streamlit. So if you're comfortable with other frameworks, use Flask, use Next is so on and so forth. Deploy apps. Right. And again stay updated. So I'll show you how to stay updated and what are the other tools that. I'll show you how to stay updated in this particular in our next slide and don't stop experimenting, right. And one thing that I have noticed from my personal experience is that AI is changing every day, right? AI what is very popular or very hot one month back. Is no longer something exciting, right? So when chat PDF or something like PDF chatting option came by three months back? It was a very hot thing, right? So this particular application hit the roof right with thousands, thousands and thousands of different paid users, right? So now this particular PDF application can be generated within like 15 lines of code or even less.

Staying up-to-date (31:26)

So some there are some libraries that will help you do it within 5 lines of code. So AI sector is moving very fast, right? So if you want to have footing in this, you need to stay updated with this and keep experimenting. So one thing is to read XYZ news today and one thing is to read ABC News. Other day, right. So if you're not doing anything productive with this news, then it's just a piece of information, not something that. Is that is usable for you? So don't stop experimenting. You can experiment in any capacity, right? So one way you can do is to develop your own applications. Let's say you don't even want to get into coding. That is not even your intent in first place. No problem, why don't you explore different tools? Why don't you explore? Make journey? Try to you. Upskill yourself in terms of how you can generate prompts so you start GPT right ChatGPT has a lot more features than. It started out with right? So if you look at Chat GPD, there were like around 5 to 6 main features that they have launched in the last 3-4 months, right? So understand how to use plug insurance plugins in chat GPD. So they have just 2-3 days back they have introduced something called custom instructions within chat bot. Understand what is happening with that. Right, and try to see how different apps are comparing, right? Let's say compare how GPT is working with bot. Right, do some experiments. Right By experiments, again, like I'm not telling you to develop. If you don't develop, you're doomed. So that is not the message that I want to confirm like convey here. So you can be anything depending upon your what your intent is or something. Just keep experimenting. So just to iterate, you can do anything, explore tools, use plugins, use custom instructions, try using different chat models, try to compare them, how one is better than the other, so on and so forth. Right. So this is the basic road map or something like that. Look at the time. OK was there a question in the chat? Some of my a learnings. OK, great. OK, what is the road map? I think this is clear, I guess. So I think I'll get back to this part at the end. So how do I like? Most of the code that I write is written by A and these are the different tools that I use,, line chain mode and GitHub copilot. I'll talk more about this at the end. So how do I stay updated? Right, so since updates are happening on a daily basis, so I follow close to 100 people or 100 accounts on different platforms like YouTube, Twitter, Discord, LinkedIn, so on and so forth to get my basic moves right. To do so, one thing that you need to probably do is I can send you the list of these 100 people and you can start following them, but that is of no avail right? So one thing that you need to do is you actively search for something that is working for you. A YouTube channel. It might work very well for me, but you might not connect to the person who is actually teaching that, right? You might like someone's a person LinkedIn, which I might not find it very useful, right? So everyone will have their own set of people are accounts that they can ideally follow, right? So be it on any social platform. So what I want you guys to do is actively take a step. Whenever you are using YouTube, whenever you're using Twitter, Discord, or any platforms for that matter, actively look for the people are accounts who give you information that is useful to you. Right. So and over the time I have, I follow more than 100 people, but I have followed these people in the course of two to three months. So every time I see a good poster something, I just go to their profile, see their previous post. If I find it useful, I'll then follow it right and I'll just bookmark that particular post for others, like for another time, so on and so forth, right? So if you want to stay updated, you might be using like at least few of these platforms yourself, right? So it can be either Reddit, Telegram, whatever, whatever that you're comfortable with, so actively identify. The resources where you can get information that is useful to you. So, but otherwise also I'll just try to compile a list of my top 15 or top 20 and try to send the mail. And another thing is, I don't subscribe to any newsletters except for one or two that I might have subscribed accidentally or something. The reason I don't subscribe to newsletters is because newsletters are generally created to. They have to send a newsletter daily or they have to send a newsletter every fortnight or something. So it's too much of information, right? So if I'm getting. Information like if I'm getting 2030 pointers every day then there is nowhere I can use this usefully right? So I generally refrain from newsletters because I just need enough information. I just I don't need more than three to four or three to five new things to see every day. So there is no way I am able to like my brain can digest 20 to 30 things every day, so I don't follow newsletters. If you find something that works very well for you, go ahead and do that. So I think this is my basic session. So what I wanted to do here is that I just wanted to talk about the master class. So I'll just take 5 minutes for this then I will open the floor for questions, right?

Masterclass (37:26)

OK, so I think the session that we are doing. So yeah. So we are doing a session on 26th August, Saturday at 11:00 AM. So it will be a three hour master class on how to build chatbot on your data, right? So what you can learn to do from this is that you will try to like you can upload any PDF and start chatting with it. So as I've already informed you how important line chain is and you can do apps without line chain as well. But for someone who is a beginner, line chain is an indispensable tool right. So we will learn basics of line chain. We'll also try to understand how chat bots are generally created. Right. So this thing will be covered and I'll also cover how to use vector databases. So in general the calls like these, if it is one hour long, there is only little bit that I can cover. So three hour workshops are where I can go through the length and breadth of the topics and try to show you how things work everywhere. So I'll show you a quick demo of what is it that we'll be building in this particular workshop so that you can understand it. So again, this is one app that I've developed maybe a few weeks back. So I'm just reloading it. So what you can do is you can upload any PDF right from this. So I'll just say that.

Diy Indian Constitution Chatbot: Tools, Strategies, And Discussions

We will DIY an Indian Constitution-Like AI (39:03)

I have downloaded a Indian Constitution small PDF. So I'll just upload Indian constitution PDF to this. So and I can ask. What are my fundamental rights?

How to Build This Particular Chatbot (39:36)

So it is giving answer from the Indian Constitution PDF that I've applied, right. So you can just go ahead and ask more follow up questions, so on and so forth. So this is what we'll be building. This is what you'll learn to build in this particular master class, right. And even if you don't know coding, that is fine. So even if you don't understand Python fully, which is totally fine. But again as always you don't need to understand 100% of things. So I can assure you that if you go through this particular session, I'll do my best to you know, explain at least 60 to 70% of whatever is being taught and the rest 30 to 40% is your own effort. As with anything in learning part, right. So this is what we will do. So I think since this will be a three hour session, we will be having a very nominal fee for it and everything will be available to you at the cost of medium pizza, right. So again what are the things that will be covered. So it will be a three hour session. We'll I'll just provide you with all the notes that is required. We'll have a follow up doubt solving session. So we will deploy the bot as well and I have shown you all the other apps that I've created so you'll get. Access to all of them. You'll get the full source code for it, and you will also have some NX builder tools, so I'll maybe show you in this session as well. So how can you code faster, right? So even if you don't like. Just starting with code, can you use any tool to help you with getting started with the templates or something? Right. So these are the things, these are some of the benefits that you can get from this and all this will be available at a cost of 499, right. So yeah, so I think this will be a three hour session and I think if you are just starting out, I think this will be a great. Thing for it. So this is the link that you can use to sign up for this session. Yeah, I think this is all that I had and one special offer I can say is that if you register today itself, you can just get the course for 399. So it's just a cost of a cappuccino or maybe a small medium pizza at Domino's. So just at this cost you can get the full access to three hour workshop. It is happening at 26th August 11:00 AM to 2:00 PM.

So you can, I think the chat. This link will also be available on the chat. And if you want to register it, you can just register it. It'll involve a UPI payment. Of 399 if you're registering it today, and if you're paying it anytime later so you can pay 499, double it. So I think this is what I had to cover in session today. So now I think we are open for questions. I think I'll send the link in the chat as well again. I think the link is in the chat, the registration link. You can just use it to register or you can just scan this particular link from the. From the screen directly with your phone and do it. Yeah, this is the form that you will see here. So it just basically involves. Name, phone number and just making a UPI payment. So yeah, it will take one minute. If you can do it now, you can do it later as well. Yeah. I think the floor is open for questions, I guess. I think someone raised their hand. I think they can unmute themselves and ask the question. I think, yeah, Arun Dhani, whatever your question is, you can go ahead and ask. OK. Is there anything on the chat?

Append the discussion around Hugging Face (43:46)

So are people not able to unmute themselves? Is that the issue? No, I'm able to unmute. OK, yeah. So Zapak, just one more point I wanted to ask is I've seen a lot of people downloading models on their personal systems and trying to experiment around so. I also heard that people use. Websites like Hugging face to do this so I will be confused on what are the advantages or what exactly like is it really required at this stage to go to that level to download. Models. Again, as I've covered in the session in the beginning, right if you're someone who is just starting to explore. It is not required. Got it? Got it is not required. Because again, The thing is, whenever we learn something, I think, let's say this is the learning curve. It's easy to get the bandwagon here and then go here, and just learning the open source models is somewhere here, right? So if you're wanting to directly jump to this part, of course it can be done. I'm not telling otherwise, but the best thing to start with is to develop some basic models. Just try to use GPT. The GPT APIs are also very cheap. So GPTAPIS are very cheap and I just like to code this particular fact, right? So if you want to generate the entire Harry Potter series which has around million words, it will only cost you around $2.00 to make it using GPD 3. Right, so the GPDAPIS are very cheap, so you just have to input your credit card. You'll get free credits if you're using it for the first time. You'll you'll get $5 credits for the first time, which is more than enough to last you for like 2 months if you are just exploring right. So yeah, if you're just starting, start using GPT 3 APIs and best way follow. Cool. Thank you. By the way, very insightful this session was. Thanks. Thanks a lot for that. OK. Any other question, Arun Dhani has raised the hand. I don't know if there is a question, Can we get a recorded session? Yeah, you'll get a recorded session for this. Hi this is Jess Agut here. I have a question like so. I'm currently working on an idea, everything with wireframing and everything is done. So my idea is to like create an app. The one problem that I came across with no code platform was anything that's very complex we are not able to create it. So in that case, do you suggest that I have to like? Work on Python And the other one you said Lang by line chain is a Python library only, so Lang chain is not a different language. So if you learn Python you will. Yeah, line chain is a part of Python itself. Maybe if I can put that very simply. So developing, so no code tools. My personal opinion about Noko tools is that what no code tools are doing is that there is some code actually that is being returned, right? No, code tools are just adding an abstraction layer to the code. The process is same. So to develop, let's say if you want to develop your app, the process is exactly the same. If you want to do it with code, let's say you have to do things 123.

Confidence (47:33)

So even if you're doing with more code, you will have to do the same 123 but in a different way in an abstracted format. So. If you're using code, going from 1:00 to 2:00 to 3:00 is very logical because the code will guide you. Now I have output, now I'll print the output. Now I'll render this. Very simple, right? So if you're just starting with no code, the steps are same, but there is a heavy abstraction layer added to it. So if you are very good with no code, then go ahead and build it. But if you are like someone who has limited idea about no code and limited idea about coding, it's a good idea to get into the code. But again, like I'm not a proponent of only code and this is only my personal opinion. OK. And also when it comes to SO you have just mentioned about apps, but would that apply the same for websites as well? When I mean app, I mean application, website, anything. OK and I have registered for the event that's there for on 29th of August so I wanted to on 26th. So I wanted to understand like how will this help me in working on this particular application idea that I have. So I exactly don't know what is the idea that you are working on, but one thing that I can promise is it will just so one way. So I think this is applicable for everyone, right? So if you see how things are being made then it will give you more confidence, right? So building a different chat bot, let's say building an application like this. Building an application like this might not be even remotely related to what is it that you're trying to build, right? Let's say, but. The tools that you use for this thing will be the same. The language that you will use, the editors that you will use that will remain the same, and it will also like give you an idea of. Where can I start? OK, Python is this simple, right? Python is this complicated or this is this simple so you'll get an idea to judge. As to if this is the thing that you want to use using Python or maybe some other framework or something, it'll help you in understanding how easy or hard or how complicated or easy it is. Just to give a better idea. OK, yeah, thank you so much. I think that answered all the questions and. Really insightful session. Thank you. Thank you. Thank you. So next question, I'll check the chat and questions. Can we read the recording session? OK. Anyone any questions? Else, any question? Any other questions?

Tools (50:44)

OK, maybe till people prepare their questions, let me just show you an application that I generally use for coding. So it is called So let's say let's say you don't know Python. Let's say you don't know any coding language, but you want to develop something, right? So a lot of tools can write code for you like code for you. But find this website which uses GPT under the hood and it has access to Internet as well, right? So whenever let's say if I want to create an application using streamlet something like this. So what I can do is I can just ask. Give me Python code to generate. What to do? Using extreme data or something? So basically let's say if I want to generate a UI for this and I don't want to write everything myself, so I can just give up a bare bones description of what I want. And you can see on the right hand side it is referring to all the articles in on the Internet. So while using GPT or something they are trained. They are mostly trained on this data. But let's say there is a new update to a library or if there is a tool that you know was not was released after 2021 September. GPT will not be able to give the answer, but this will give you. Right so you can use this tool like if you are someone who is starting to code or if you want to speed up your code significantly you can go ahead and use this thing So it will help you write code. And what you can also do is you can just copy paste a piece of code. Let's say you don't understand any of what is written here so just copy paste the code here. OK, sorry, just press enter otherwise. Vishal to stop generating. So let me. I'll just ask Can. Explain this code step by step. I have not used Python before. Let's say you have never used Python before and you want to ask it to explain the code, it will do a pretty decent job. So it is telling us the best of what is happening, why is it happening, so on and so forth. And again, everything takes a learning curve, but I think a tools today will help you significantly speed up the process of learning as well as developing. OK. So I think we can go back to questions now if there is. Yeah, I think if there are no questions, I think if you feel free to send your questions on the chat as well. I'm seeing the chat, so you can send the questions on chat as well. I think we are at time, but I'm on the call for the next 5 to 10 minutes if anyone wants to. Ask something, I'm still here and if anyone wants to leave the call, I think the basic part is done here. So I think once again, like there will be a session on 26th August. Which will talk about how to build chat bots on your data. So this is the link to register. And yeah, this is something that you can scan, register within a minute. So yeah, if you have not registered itself and if this is something that you are interested in, you can go ahead and do that at just ₹400 today. And I think we can. So people who want to exit the call can exit and I'm on the call for the next 5 minutes. If anyone has any specific doubts or anything that they wanted to ask, I'm available on the phone. I have AI have two questions. The first one may be very silly but I was C&C developer before so I'm very new to Python.

Understanding Ai Applications And Chatbot Setups

ACeDEv first question array similarities with other languages (55:45)

So I just tried to explore very few of Python basics and I found it similar to CC plus plus. But I wanted to take your opinion, I mean when I want to learn Python. Do I see any similarities or it's very different language? So with coding language, right? So if you're wanting to learn Python or JavaScript or anything, the only thing is if any coding language it is. You already know coding, right? So it doesn't matter if you learn C++, if you don't, if Rust or if you know let's say Java or anything. It doesn't matter which coding language you know. If you know any coding language you can easily or like even or probably you'll feel. Very liberated to learn Python because Python. So in C++ or Java or something, you have to write a lot of things. You'll have to declare a lot of variables, you'll have a lot of brackets and all. Python will be. A breeze like for Someone Like You right? So again So if you are starting on a coding journey and if any other coding language CC plus plus Java, rest go. Doesn't matter. If you know any coding language it's good to go. OK and my second question is you explain about the It is I think explaining you about the code or getting tips from the code but. What the Goa tool that we use? You have to write the code, compile or something. Or maybe I missed it, you already thought about it, so I use VS Code. OK, so we score. Yeah, I use VS Code.

Bots Work Setup (57:31)

Combined with GitHub. OK, and for coders, other tables find as a VS Code extension as well. So maybe I can just show you my work setup real quick and non coders might get intimidated, but I'll still take that risk.

Q&A (57:52)

So let's say this is 1 code that I'm developing. So OK this I don't want to disturb this code. Let's say I'm opening this. So here what you can do is just press command I so you have a you can ask for doubts to find in the box itself.

Boxing (58:09)

So this is the way I write code. So you so this is where this is where you write and if you have any doubts just ask a the question it will give you the answer just copy paste on the left and right hand side. And I'm using GitHub copilot so whenever I write something it'll give me suggestions already right So if I say. Def. Something else So it is already giving me a suggestion as to what I want. So sometimes it will give you useless suggestions but most of the time it is contextual so very useful. OK and one more follow up question. I see this is a Linux environment right? So the can I still use in Windows the Visual Studio and still have these kind of extensions or your of course of course of course. Thank you. So I think what is the easiest way to scrape a particular website? Any suggestions on it? Yeah, let us just use fine for it, right? OK, so and there are again multiple libraries, multiple ways to do this. Say I want. Scrape website, let's say using Python. Give me the code. Beautiful soup is 1 library. It's called BS 4:00 so it will probably use it. So yeah, it is telling you how to do this. So this is the example. This is how you do it. Got it. And I need to run this in my environment and OK exactly. Thanks. Thanks. Any other questions? Hello. Yeah, go ahead.

AI applications (01:00:18)

Hello Sir my name is Vinayak and I'm a fresher right now means I just joined one of company at Bangalore and I'm also a freelancer. I create website for my clients. So means currently means I don't want to create the website. You know the AI applications like you are creating means I just want to leverage this platform. To do my work fastly so I can reach more client and create more website like that. So where I can find these applications which has helped me. I think there are many coding tools, right? So there are many coding tools, many VS Code extensions that are available online. So the one that I have sent you is something that I use mostly. So you can use that. So and again, with respect to tools, resources or anything, if someone, if you are someone who is exploring it, you know let's say on a constant daily basis or something, you will find a tool that is working best for you. So I can give you names of XYZ tools like let's say find GitHub, Copilot, or maybe there are some discord bots also that you can use, so on and so forth. But what works best for you is something that you have to figure out yourself. You can just start with usingfind. com. It will give you a significant boost up right away. Yeah. Thank you. Thank you, Sir.

Lambert Model Analysis

Lambert Model (01:01:52)

I had one question I was looking at this, the new model right Lamar 2 model. I heard in some of the videos right, it's possible to download it and use it in our laptop like have you tried that one? Is it possible? Yeah, it is possible. It's totally possible. So here what happens is whenever the model is let's say in the application that I have created. Model is something that is there in open as server or something. So if I'm using my application I'll make a API call. And I'll get the answer like this right. So here what will happen is even. Whenever you're downloading something, your app will have the model along with it. So whenever you give an answer it will generate itself something like that. So again like it is totally possible to install models on your own computer and do it. But. Again, the same suggestion, if you're just starting out, I don't advise you to go down that room because there are a lot of different variables to do it, how to download, how to use it, how to do the setup. So that is very complicated for someone who is just starting out. So if you are wanting to, you know, just start out, you can just use GPT 3. Does that answer your question? There isn't. Yeah. Thanks. Any questions? Let me just check the chat. OK. I think so I think we can end the call right now. Maybe I'm just. I'll just stop for 1-2 minutes and then I'll be ending the call. OK, then great guys. So thank you all for attending your session today.

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Wisdom In a Nutshell.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.