Transformers, explained: Understand the model behind ChatGPT

Рет қаралды 19,185

Leon Petrou

Күн бұрын

Пікірлер: 88

@rajatverma4766 2 ай бұрын

I must have watched like a million videos on Transformers but this is the first time i have completely understood it.

@LeonPetrou 2 ай бұрын

So glad to hear that! Let me know what tutorial video you'd like me to make next.

@sreelakshmi7472 Күн бұрын

I'm so glad I found this video. Great explanation, Leon!

@WangHongGuang2012 2 күн бұрын

I watched several videos on this topic, this video is definitely outstanding. The explanation is clear and easy to follow. So thankful!!

@Muddzdk 2 күн бұрын

I watched many videos on the topic and this one was the easiest to follow. Visuals, animations, analogies and real world examples helps a lot, so keep using those in your videos.

@LeonPetrou 2 күн бұрын

Thank you so much for the feedback! I'll make more videos like this.

@ravindranshanmugam782 9 ай бұрын

Excellent, went thro' multiple videos on basic understanding of Transformers. This is the best one I could quickly grasp. Effortlessly explained, Well done !!

@LeonPetrou 9 ай бұрын

Thank you Ravindran! I try my best to teach things the same way that I'd like to be taught, which is simple and step-by-step. Let me know what other videos you'd like to see from my channel.

@ravindranshanmugam782 9 ай бұрын

Hi Leon, it would be great if you can make videos on Langchain and its application which are trending now. You can also add topics like Vectordatabase, Embedding, word2vec and so on. Anything on GenAI is hot now in tech space. Thanks.

@ovidioe.cabeza4750 7 ай бұрын

Same for me, I am a python backend dev and getting transformer was being tough, but you helped me a lot, thank you!

@Yaser-z9j 6 ай бұрын

Me too@@ravindranshanmugam782

@NithishAnuth 5 күн бұрын

Went through many video on transformer but this one THE BEST !!

@LeonPetrou 4 күн бұрын

Appreciate that! What video should I make next?

@nimitbhandari2859 2 ай бұрын

Finally a simple and terse explanation for transformers, loved it! ❤😊

@wp1300 8 ай бұрын

1:12 ANN 2:26 GPT-1 ~ GPT-4 3:34 LLM 7:09 Transformer architecture 7:45 Tokenization & Detokenization 8:17 Step 1 10:14 Step 2 10:20 Token embeddings 14:48 Step 3 15:10 Position Enbedding 16:58 Step 4 17:17 Self-Attention 18:52 Multi-headed self-attention 19:55 Step 5 20:27 Feed-Forward 22:02 Step 6 22:32 Step 6

@sandromartins3579 2 ай бұрын

Amazing explanation! Thank you!

@antarikshshreshthi 5 ай бұрын

I believe that this is the best video for transformers, embeddings and tokenisation on the internet!!!

@LeonPetrou 5 ай бұрын

Appreciate that! Let me know what tutorial you want to see next!

@atharvadeshpande4976 29 күн бұрын

Thank you for the explanation. Too good.

@60pluscrazy Ай бұрын

Best explanation 🎉🎉🎉

@fhs14647 23 күн бұрын

very good explained, thank you so much!

@michaelzap8528 7 ай бұрын

best. Finally i understand how gpt work now. Thanks male, u the champion.

@marcinnnWL 4 ай бұрын

Best material for me - easy explanation in my way of thinking. Thanks Leon. :)

@LeonPetrou 4 ай бұрын

Appreciate that!

@sukumarane2302 3 ай бұрын

You made it simple … really excellent!. Thanks 🙏

@monicabhogal 2 ай бұрын

Awesome , beautifully explained ❤

@rajathslr 5 ай бұрын

You have put a lot of effort for this wonderful video, thank you so much

@ravideepa 4 ай бұрын

Excellent tutorial on this concept. Awesome 👏

@vj7668 7 ай бұрын

Excellent !!! Thanks for simplifying it. Loved it !

@LeonPetrou 7 ай бұрын

Appreciate that, thank you!

@KonstantinosEvangelides 6 ай бұрын

Can you do a separate video exploring further what are embeddings and what does the vector embeddings represent more thoroughly. Great video!!

@LeonPetrou 6 ай бұрын

Great idea! I'll do this next.

@sudhanshusaxena8134 8 ай бұрын

Great explanation.

@LeonPetrou 8 ай бұрын

Thank you very much!

@anibeto7 9 ай бұрын

It was indeed a very informative video. It cleared a lot of the important ideas. Thanks a lot.

@DarrabEducation 4 ай бұрын

It's amazing, more such videos will make you in the top with another video as an example.

@F30-Jet 5 ай бұрын

6:56 to add more clarity on Pretrained; Pretrained means the model has acquired general-purpose knowledge.

@programminglover2976 7 ай бұрын

thank you so much.. really reallly well explained.

@JRKyt00 6 ай бұрын

Agreed--best explanation I've found. Now I get it (well...)!

@MotulzAnto 8 ай бұрын

THANK YOU! easy explanation..

@LeonPetrou 8 ай бұрын

Appreciate it!

@crazyant1080 4 ай бұрын

Thanks a lot.

@PigiMontieri 2 ай бұрын

Wow thanks ♥️

@karannesh7700 7 ай бұрын

thx for this great video !

@LeonPetrou 7 ай бұрын

Appreciate it!

@JohnCohen-ur5hk 7 ай бұрын

Very Good Explanation. Thank You

@Clammer999 8 ай бұрын

Wow, this is one of the easiest to understand video on how transformers work. You also explained very tokens and embeddings which I was searching for. I’m a complete newbie and I kept hearing nuerons and neural networks. Is a neuron a physical device/hardware or it actually an algorithm? And a neural network is not a physical network?

@LeonPetrou 8 ай бұрын

Thank you! Neural networks, and everything explained in this video is all software (except biological neurons which is in a human brain), it is all algorithms. It's basically just code. The hardware that the code runs on usually just requires high processing power / RAM. This can be a CPU or GPU.

@otenyop 5 ай бұрын

Great explanation

@samrmit6253 5 ай бұрын

brilliant video thanks

@Omniassassin7 10 ай бұрын

This is amazing, thanks a lot man! Quick question, how are the self-attention layers produced? Does the model dynamically “decide” which contextual layer to use depending on the prompt, or is the set of layers learnt during training?

@LeonPetrou 10 ай бұрын

My pleasure man, glad you like it. That's a great question. The structure and behavior of these self-attention layers are determined during the model's training phase, not during inference. Simply put, the model learns which words in a sentence should pay attention to which other words to better understand the sentence's meaning. This learning process is fixed once the model is fully trained.. it does not change or decide on a different structure when it's given new prompts to process.

@changliu7553 4 ай бұрын

@@LeonPetrou Thanks. I am starting to think the attention could be used for google search - this way we don't have to use stupid SEO's. The same question can be asked 1000 different times.

@Yaser-z9j 6 ай бұрын

Awesome 👌 thank you so much, You are amazing

@baigsaab47 4 ай бұрын

Hi Leon. Can you kindly make a video explaining LLM's library vLLM.

@poorjahangiri11 Ай бұрын

vey well done!

@LeonPetrou Ай бұрын

appreciate it!

@AvaMichl 5 ай бұрын

Does one word always equal one token embedding?

@LeonPetrou 5 ай бұрын

@@AvaMichl no not always, it’s just an simple way to think about it, but on average, one token is 4 characters of text.

@GilCohen-z5t 4 ай бұрын

It would be great to get the slides :)

@LeonPetrou 4 ай бұрын

Sure, why would you like the slides?

@GilCohen-z5t 4 ай бұрын

@@LeonPetrou I'm giving a talk at my small data startup on GenAi. I was hoping to incorporate some of your fantastic work from this video into my presentation. I’ll be sure to give you full credit and direct people to your video.

@LeonPetrou 4 ай бұрын

@@GilCohen-z5t Sure, happy to share the slides with you. I'd appreciate the traffic to the video. What email would you like me to send the slides to?

@prathamsinghjamwal4725 Ай бұрын

Sir could you also provide the pdf of this video

@kamal9991999 7 ай бұрын

This video is a lot better one ☝️

@LeonPetrou 7 ай бұрын

Appreciate that!

@rhktech 7 ай бұрын

very well explained (Y)

@changliu7553 4 ай бұрын

After watching a bunch of videos, I think yours clarifies many things! Thank you. Question here. You lost me between "the fisherman caught the fish with the net" and "the cat is sleeping". They have connection? If you are trying to translate to another language, I can understand. But why does the GPT say "The cat is ...."? what was the input there? Thanks

@LeonPetrou 4 ай бұрын

In that example "The cat is" is the input/prompt.

@abooaw4588 9 ай бұрын

Bravo 🇨🇵Dommage que ce très bon niveau de d'explication n'est réservé que pour nous qui comprenons l'anglais. Lecun et Bengio en sont pour beaucoup. Heureusement que le nutshell n'est pas traduit par GPT à la noix!

@LeonPetrou 9 ай бұрын

Merci beaucoup for your thoughtful comment! I'm glad you found the video informative. Your point about language accessibility is very important to us. We're actively exploring options to include subtitles in multiple languages in our future videos to ensure more viewers can benefit from our content.