Відео

This is actually incredibly helpful. As someone who is currently working on building an LLM Chatbot for his university, I needed a detailed and cohesive explanation on the underlying infrastructure as well as the code logic behind building different components of a Large Language Model. Thank you so much mister.

Where can I find the notes?

Thanks a lot .

Question: could any open source llm model be able to substitute chatGPT 3.5 turbo? I tried google/flan-t5-large but it didn't work. Thanks

Awsome video, thank you so much for sharing all these valuable knowledge! thanks for the excellence contribution to this learning communities!

accuracy formula is wrong in notes kindly check 2 times FP you mentioned instead of FP+FN in denominator small typo

This problem is one of why we not have true AI now, if your method can fix some so..

I agree with you 100%

Very helpful sir, could visualize the whole computation. Although sir, is attention score to relate with other words in the sentence or just find the distance between them in the sentence?

Waiting for transformers explanation in same manner. Great work broh

Very good - very useful - the best explanation 👍

Great Explanation bro

I don't know why I am seeing posts related to you on LinkedIn. 😂 You are doing a really good job 💯 Thanks for sharing knowledge for free that many people of your caliber either don't share out of their ego. ❤

Where are you from

Hello sir, can you teach reinforcement learning, markov decision process etc. from scratch. Thank you sir!!!

"I look forward to watching your videos daily and truly enjoy your content. To reach a broader audience, could you consider adding audio tracks in multiple languages, especially Hindi 🇮🇳? It would make a big difference! I'll keep you in my prayers at the temple for your continued success. Your work is unique and inspiring

Key Takeaways: + Have a Blended approach ==> Solid Conceptual knowledge along with Practical approach + Follow series along with making Notes Thank you Team Vizuara 😍

Hi, great lecture but how we can use in the organization perspective because these tools are publicily available , what if we want to make specifically for the organization and how it can be powerful then RAG

Very nice

I have already

Great work!

Great work!

i'm confuse while doing transpose in every step .

Key Takeaways: Chain Rule & its applications + If "f" is a function in terms of another function "g" where function "g" is defined in terms of a variable "x" Then, change of f wrt x is calculated through change of f wrt g multiplied by change of g wrt x i.e we will be finding first how 'f' changes wrt 'g' & then find how 'g' changes wrt 'x' + Application of Chain rule in ML: --> Chain rule is fundamental in back-propagation --> Neural network's loss functions are complex function where we have to minimize & Chain rule plays a key role in this

hey i am in 1.1k people visited this masterpiece.

Amazing job!!!

Key Takeaways: + This lecture covers Basic Concepts of Calculus + Calculus is a branch of Mathematics that focus on the rate of accumulation (Integral Calculus) & rate of change (Differential Calculus) + It is Core to Optimization in ML Algorithms --> Gradient Descent, Loss Function + Limits & Continuity + Limit is the value of a function as the input approaches a certain point + LHopitals Rule + Derivative of f(x) is the rate of change of 'f' w.r.t 'x' + Differentiation Rules --> Power Rule --> Sum Rule --> Constant Multiplication Rule --> Product Rule --> Quotient Rule --> Chain Rule + Applications of Derivatives --> Finding Slopes and Tangents --> Optimization

Awesome so far. Thank you for your efforts.

Is the mentioned course link in your description different from dr raj dandekar's NNFS course?

@Vizuara Is it possible to share the whiteboard notes you write as pdf ?

Amazing playlist sir

Please also provide the link of the book also . Your work is truly amazing , Thank You

Very well done for so such a caring attitude to convey the main message of TGN, this is what is called dedication.

We need English version plzzzzz

I thought k-fold is only conducted on training datasets not the entire dataset?

Question : You mentioned Sentiment Analysis as one of the things an LLM can do. The problem with that claim implies that there is a key-value pair cached somewhere in the LLM’s memory or training (like in VADER/polarity based sentiment analysis) which can be used to rank every token the LLM has been trained on with a numeric score, in order to generate the overall sentiment score for a given word, phrase,sentence or paragraph. I don’t believe this is the case.

where can we download the notes?

Thanks alot. I have not seen anyone teach ML this way. You took it to the grassroot.

Its good because if all go to stay there but why don’t they think they should try to improve the conditions of own country though difficulties are more here but see people like Ashawini Vaishnav well educated & our ISRO Scientists.

video stops playing at 12:26, thanks

Your didactic skills are stellar. Very impressive.

where is the classification.ipynb file shared?

i want to ask where does the decoder in transformer receive partial output text from from i.e this is an __ ( in german)

Really awesome content . I am going back to my College days when my friends would make note and teach us before exams . Thank you buddy .

wowww amazing

Amazing initIative. Can we please get the link of the notes you have created? or a pdf?

Wonderful, professional, and highly informative! Thanks.