Starting Generative AI On Cloud New Series- AWS And Azure

Summary

YouTuber Krishak announces new video series on deploying generative AI applications on cloud platforms like Azure and AWS. He plans to cover Azure AI Services and AWS Sagemaker, including fine-tuning, deployment, and production. Krishak emphasizes the importance of having Azure and AWS accounts for practical experience. He outlines prerequisites like NLP knowledge, Python, and familiarity with frameworks like Langchain and Hugging Face. He also stresses the value of dedication and building an online presence to showcase skills and projects.

Takeaways

• 😀 The speaker, Krishak, is launching new YouTube playlists focused on implementing generative AI applications in cloud platforms like Azure and AWS.
• 🌐 The playlists will cover Azure AI Services and AWS Sagemaker, including topics like fine-tuning models, deployment, and setting up endpoints.
• 🔧 Krishak emphasizes the importance of understanding LLM (Large Language Models) Ops, which is an extension of MLOps tailored for language models.
• 💼 The content is aimed at helping viewers make career transitions and meet industry requirements for working with generative AI in the cloud.
• 💡 Krishak will demonstrate how to take AI models from development to production, highlighting the significance of this skill in the job market.
• 💼 The speaker suggests that having an Azure and AWS account is essential, even for those with limited budgets, as it provides hands-on experience with cloud services.
• 📚 Prerequisites for the playlists include knowledge of NLP, Python programming, and familiarity with frameworks like Langchain, Lama, and Hugging Face.
• 💼 Krishak advises viewers to be dedicated and build a strong online presence to showcase their skills and attract potential job opportunities.
• 📈 The playlists will be progressive, starting with basics and building up to more complex applications, aligning with Krishak's teaching philosophy.
• 🔗 The speaker encourages viewers to support the channel, subscribe, and turn on notifications to stay updated with the new content.

Q & A

What is the main focus of Krishak's YouTube channel?

Krishak's YouTube channel focuses on Generative AI, showcasing the use of various frameworks and Foundation models for creating end-to-end applications.

What new series is Krishak planning to start on his channel?

Krishak is planning to start a series on implementing Generative AI applications in Cloud platforms such as Azure and AWS.

What specific services within Azure AI Services will Krishak be covering?

Krishak will be covering Azure Open AI Services, including applications related to images and text.

How does Krishak plan to demonstrate the use of AWS SageMaker?

Krishak plans to demonstrate how to create end-to-end Generative AI applications using AWS SageMaker, including fine-tuning, deployment, and setting up endpoints.

What is the importance of fine-tuning and deployment in the context of Krishak's videos?

Fine-tuning and deployment are important as they cover the process of taking AI models from development to production, which is a critical skill in the industry.

What does Krishak mean by 'LLM Ops'?

Krishak refers to 'LLM Ops' (Large Language Model Operations), which is an extension of MLOps specifically focused on the operations and management of large language models.

What are the prerequisites Krishak mentions for his new series?

The prerequisites include knowledge of NLP, Python programming, understanding of machine learning and deep learning concepts like word embeddings, and hands-on experience with frameworks like Hugging Face.

Why does Krishak emphasize the need for an Azure and AWS account for his new series?

Krishak emphasizes the need for Azure and AWS accounts because the new series will focus on cloud-based implementations, and having these accounts allows viewers to practice and gain practical experience.

What advice does Krishak give regarding dedication and learning from his content?

Krishak advises viewers to be dedicated, focused, and to build solutions and projects to showcase their skills and credibility online.

How does Krishak plan to manage costs for viewers who might be concerned about using Azure and AWS services?

Krishak plans to demonstrate services that are cost-effective, using smaller datasets for fine-tuning to minimize charges, and ensuring to stop services to avoid unnecessary costs.

What is Krishak's final request to his viewers at the end of the script?

Krishak's final request is for viewers to support his channel by subscribing, pressing the Bell notification icon, and looking forward to the next video.

Outlines

00:00

🌟 Introduction to Generative AI and Upcoming Series

Krishak introduces himself and his YouTube channel, which has been focusing on generative AI for the past few months. He has used various frameworks and Foundation models to demonstrate how to create end-to-end applications. He expresses gratitude for the positive feedback he's received and announces two new series: one on Azure AI Services, exploring image and text applications, and the other on AWS Sagemaker, focusing on end-to-end generative AI application creation, including fine-tuning, deployment, and production. He emphasizes the importance of covering fine-tuning and deployment in the production phase, which is crucial for industry applications. He also mentions the upcoming inclusion of Google Cloud and Google Vertex AI in future videos.

05:02

💼 Prerequisites and Encouragement for Learning Generative AI

Krishak outlines the prerequisites for his upcoming series, emphasizing the need for a strong foundation in NLP, Python programming, and understanding of machine learning, particularly deep learning and concepts like word embeddings and transformers. He suggests having hands-on experience with platforms like Hugging Face and common web frameworks. He also stresses the importance of having Azure and AWS accounts, even with a small budget, to gain practical experience with cloud services. He encourages dedication and continuous learning, highlighting the significance of building a strong online presence and showcasing one's skills. He ends with a request for viewers to support his channel by subscribing and engaging with the content.

Keywords

💡Generative AI

Generative AI refers to artificial intelligence systems that can create new content such as text, images, music, and videos. In the video, the creator discusses his experience with various generative AI frameworks and models, indicating that this technology is central to the content he produces. He also mentions creating end-to-end applications with these models, highlighting the practical applications of generative AI.

💡Foundation Models

Foundation models are large-scale machine learning models that are pre-trained on a wide range of data and can be fine-tuned for specific tasks. The script mentions using multiple foundation models, suggesting that these models form the basis for the generative AI applications discussed in the video.

💡Fine-tuning

Fine-tuning involves adjusting a pre-trained model to a specific task using a custom dataset. The video script discusses fine-tuning foundation models with custom data sets, which is a key process in tailoring generative AI applications to particular use cases.

💡Cloud Platforms

Cloud platforms like Azure, AWS, and Google Cloud provide services and infrastructure over the internet. The video mentions creating playlists on implementing generative AI applications in these platforms, indicating that the cloud is a significant environment for deploying and managing AI applications.

💡Azure AI Services

Azure AI Services are a set of tools and services provided by Microsoft Azure for building AI solutions. The script specifically mentions creating a playlist on Azure AI services, suggesting that these services will be used to demonstrate how to develop and deploy AI applications in the cloud.

💡AWS SageMaker

AWS SageMaker is a fully managed service that provides developers and data scientists the ability to build, train, and deploy machine learning models quickly. The video script discusses creating a playlist focused on using SageMaker to create end-to-end generative AI applications, emphasizing its importance in the cloud-based AI development process.

💡Deployment

Deployment in the context of AI refers to the process of making a machine learning model available for use in a production environment. The video script mentions covering deployment extensively, including taking models to the production phase, which is crucial for real-world application of AI technologies.

💡LLM Ops

LLM Ops, a play on ML Ops, refers to the practices for managing and operating large language models (LLMs) throughout their lifecycle. The video script mentions LLM Ops as a key topic, indicating that the creator will cover not just development but also the operational aspects of maintaining and scaling LLMs.

💡Prerequisites

Prerequisites are the prior knowledge or skills required to understand or engage with a particular subject. The video script lists several prerequisites for the audience, such as knowledge of NLP and Python, emphasizing the importance of foundational skills for the practical application of AI.

💡Hands-on Experience

Hands-on experience implies practical, direct involvement in a task or activity. The video script encourages the audience to have hands-on experience with frameworks like Lang chain and Hugging Face, suggesting that practical application of theory is key to mastering generative AI.

💡Online Credibility

Online credibility refers to the reputation and trustworthiness one establishes on the internet, often through sharing knowledge and work. The video script stresses the importance of building online credibility by sharing projects and solutions, which can lead to job opportunities and recognition in the field of AI.

Highlights

Introduction to a new series on generative AI applications in cloud platforms.

Demonstrations of using different generative AI frameworks like Lang chain, Lama, and Foundation models.

Showcasing how to fine-tune foundation models with custom datasets.

Success stories of viewers making career transitions with the help of the channel's content.

Announcement of two new playlists focusing on Azure AI Services and AWS SageMaker.

Explanation of the scope of Azure AI Services in relation to OpenAI models.

Details on the AWS SageMaker platform and its support for various foundation models.

Coverage of fine-tuning, deployment, and production phase techniques for generative AI applications.

Introduction to the concept of LLM Ops as a specialization within ML Ops.

The importance of having an Azure and AWS account for practical implementation.

Strategies for managing costs while using cloud services for development.

Prerequisites for the new series, including knowledge of NLP, Python, and experience with web frameworks.

Emphasis on the necessity of hands-on experience with Hugging Face and understanding of word embeddings.

Advice on building a strong online presence and credibility through project sharing.

Encouragement for viewers to be dedicated and proactive in their learning journey.

A call to action for viewers to subscribe and support the channel for continued content creation.

Transcripts

hello all my name is krishak and welcome

to my YouTube channel so guys from past

two to three months I've been uploading

a lot of videos specifically related to

generative Ai and there I've actually

used different different Frameworks like

Lang chain Lama index use multiple

different Foundation models whenever a

New Foundation model is probably coming

up I'm actually also showing your demo

how you can actually use it for creating

an endtoend applications you know and

with respect to all these kind of

foundation models I've shown how you can

also fine tune with your own custom data

set and many more things now uh based on

all these videos I've seen like many

many people have also made successful

career transition I've been getting a

lot of messages and they are really

appreciating the specific playlist now

considering this I had a conversation

with them and I asked like what more

things are specifically required in the

industry or what all work they are

actually doing then they had actually

requested for creating or a playlist

specifically on implementing generative

AI apps or applications in Cloud

platforms Cloud platform basically means

uh in Azure in AWS or in Google Cloud so

considering this I'm going to probably

start this specific series uh uh I'm

going to start two parallel series two

specific playlists one is in Azure AI

services or specifically if I talk about

Azure open AI Services I'm going to use

over there and what all things are there

with with respect to images with respect

to text all those kind of applications

I'm going to develop through that the

second one is probably working with uh

AWS Sage maker and trying to show you

how you can actually create an end to

endend geni application uh it will be

including fine-tuning it will be

including deployment it will be

including endpoints and many more things

one very important thing that I'm

actually going to cover in all the

specific playlist is about how you can

actually do the fin tuning along with

that how you can take it to the

deployment part till the production

phase how you can do the entire

deployments how you can come up with

this entire amazing techniques of llm

plat llm Ops okay so it was mlops before

now llm Ops specifically where we will

be working with this kind of llm models

itself uh if I compare Azure and awsc

since you know that uh Microsoft is

investing a huge amount of money in open

AI right so definitely all the services

that are specifically available with

respect to opena will also be available

in this Microsoft azure so most of the

models that we'll be seeing will be uh

related to uh this openi itself in the

Azure Cloud itself but if I consider AWS

like you have AWS Sage maker you have um

uh in AWS Sage maker specifically if I

talk about there are different different

Foundation models both open source both

clo play paid Source everything those

kind of models are also there right

mostly they charge with respect to the

inferencing and with respect to the paid

models they may be charging based on the

number of requests

right so I will be covering both these

things and later on once I probably

upload some n number of videos later on

we'll also be seeing how we can use uh

Google cloud or how we can use Google

Vortex for creating all this kind of gen

application where we will try to also

use Google gini pro 1.5 right so all

these models obviously till now I have

shown you everything with respect to

development like how we can develop it

in a coding environment itself but again

my main aim is to take it to the

production and that is super important

because in companies when you probably

go they definitely will be requiring all

those things to you will be requiring

all those things to impress the

interviewers right so all these

techniques will probably be getting

covered okay now one of the common

question that I usually get whenever I

start any series any series of playlist

right like what are the prerequisites

see guys with respect to prerequisite

obviously knowledge of NLP is required

right you really need to have a good

knowledge of NLP because here it is more

about implementation thing but

theoretical implementation theoretical

understanding if I say word embedding if

I probably say vectors you should be

able to understand all those things

right so all those videos and playlist

have already created so that Foundation

is definitely required first is your

Python programming language NLP in

machine learning NLP in deep learning

you should really know how does a

transform work what is word embeddings

what are the different kind of word

embeddings at least some hands-on

experience with hugging face at least

some hands-on experience with some of

the common web Frameworks like Lang

chain Lama index and lot right so if you

are following my playlist if you are

following me if you are following my

content I think you should never have a

problem with respect to that because I

always make sure that I plan my content

in a specific way that initially all the

basic things is basically shown it to

you and later on we keep on building

amazing applications as we go ahead

right so that is the first prerequisite

you really need to have right all the

necessary things before we start the

specific thing the second thing that I

really want as a prerequisite is that

see guys um you need to have Azure

account and AWS account I know people

will be saying chish we don't have

credit card we don't have this we don't

have that try to try to see it guys

because see um many many people right

will try to take this as an experience

if you have worked little bit or not I'm

not saying that invest $100 $200 every

month no right hardly $10 I will try to

create a videos in such a way that both

in Azure account or in AWS account

hardly with respect to $10 we will go

with the pay as you go Services we I

will try to show you all the services

that is probably required and maximum

amount of money will just go with

respect to fine tuning and for that also

I will take care that I'll take a

smaller data set so that much charges

will not happen from your side right and

at the end of the day every video that I

will be creating I'll be making sure

that I'll stop that services in front of

you right so it is important you really

need to have an AWS account and Azure

account because I'm telling you guys if

you sincerely follow this and nowadays

companies are are wanting the skills

generative AI something or AI something

that will probably be included in every

text tack you cannot run away out of it

you know tomorrow anywhere that you

probably go every text tack that is

involved you have to work generative AI

or AI skill sets is going to come over

there right so it is good that we

basically Bally know with respect to all

the cloud platforms also see AWS Azure

everywhere generative a AI something is

basically included over there right so

it is important you really need to have

these two specific accounts without

these two accounts you cannot probably

go and even check it out because all

because we will be developing everything

in the cloud platform itself right so

this is the second prerequisite the

third prequisite is that if you are

learning right please make sure that you

be dedicated in it right when in terms

of implementation in terms of sharing

knowledge in terms of Building Solutions

in terms of building projects and share

it with the entire world your online

credibility is very much important you

know because people will come and ask

you like how you can actually do it

companies can come and give you some

freelancing work many people will come

and refer you right unless and until

they don't see the talent and in this

age where you have an internet right I

think there is a huge scope of

showcasing your talent throughout the

world right and that is important right

right I've seen many people hey Krish

there is no jobs with respect to data

science some people are able to crack

three to four jobs how they able to do

it because they keep on working on it

right so just don't keep on complaining

instead it is your life it is you who

has to probably take it forward so it is

you you basically who has to work hard

so definitely keep on working hard keep

on doing well with respect to this

unless and until you don't devote your

100% or 120% trust me it is not going to

work out so this is a request from my

side any playlist that you probably

follow please focus and learn it in a

separate way right a way that can you

can actually create a differentiated

skill right and just don't follow my

playlist only see after completing my

playlist always keep on looking for more

and more examples build more solutions

right so yes uh this was it for my side

one last request guys as you all know

lot of efforts is there in developing

all this kind of videos and projects so

please make sure that you keep on

supporting please do make sure that you

subscribe the channel channel uh press

the Bell notification icon and yes I

will see you all in the next video have

a great day thank you and all take care

bye-bye