Thanks Umar😊

Thanks for great video. Shouldn’t we use “residue=x.clone()” at 1:22:30 for residual connection. Otherwise residue variable will get updated also

For the first minibatch in off-policy learning, the ratio of offline and online log probas would be 1, right? It's only after the first minibatch pass, online policy would start producing different log probas for action tokens

Thanks!

"This explaination is all I need" to crack my DS interviews. I really liked the smooth buildup of Transformer concepts in this video. Thanks a lot man!!

One random afternoon last year I decided to watch the whole video, and now I have my own LLM with 1B parameter with your code. Thank you so much. Don't ever stop inspring new ai programmers! Greetings from Philippines.

Sir, I understood most of the content, but I'm struggling to grasp how the decoder block of UNET was designed. I tried to understand it and even attempted to write it on my own, but I couldn't manage. Could you please guide me on where I should focus or share any resources that can help me understand this better?

Excellent video! Could you provide the notes, I would greatly appreciate it.❤

next explain vision transformer 🙏

Please make videolllama explanation in the same way❤❤❤

❤

Great, as always, thank you sir!

Hello Umar, i'm always be amazed with your videos. Can you make a video about how to run huggingface models on Google's TPU's. It's hard to understand torch_xla and documentation is not quite well. When working with billion parameter models like LLaMA 7b etc. it would be easier to run it on Colab's TPU's. Also Google has a program calling the TPU Research Cloud(TRC) and it gives us researchers a great opportunity about training or fine-tuning this billion parameter models. It would be great if there a tutorial about how to utilize the TPU's. Have a great day!

What a great contribution Mr Umar …. Deep Respect for your work ❤

thank you so much, great work

bro build GPT backend as a training course

worth every second

Completed it today. One word `Amazing`. Looking forward to the flash attention with Triton.

Thank you so much.... Best explanation on yt i could find on this topic....🙌

Thanks!

Best explanation of transformers on youtube!!

Thanks!

Great video Umar, thank you so much. Any chance to you release the finetuning video to guide how to fit this model to my own dataset?

3:39:42 哈哈哈 真可爱

This is ABSOLUTELY AMAZING!

The best video for learning RLHF.

Thank you! It has been a life saver :) It is definitely the best explanation of such important concepts by a long way! Thank you so much.

The best video explanation I have ever got on Mistral!! Thank You so much for your efforts.

Thanks!

AI圣经

Great !!! Molto grazie !!!

You are my lifesaver thank you very much

thanks for the masterpiece, can you make a video to talk about the recent famous model: Recurrent Memory Transformers (RMT), and how to make this new model compatible with transformers in HF.

How can i train a transformer model from scractch for pdf summerization.... Is that the same procedure...???? Can you please reply. I'm a student who knows only fundamentals. And wish to knows more..

How can i start to deep dive into the RAG? I have studied software engineering in the university as bsc and these days i’m studying E-Commerce as msc, which is mixed up completely with Ai.. and i exactly want to continue to RAG🥺

thanks

If you only have time to watch one video about flash attention, this one is the one.

Actually watching it for the second time now, because there are so much valuable informations in here :D

Great video, thank you very much for this!

Brillant video on transformers with key math explanation.

thanktha

So it's the cat that is behind the scene kzbin.info/www/bejne/f4SxlYSZhc2mqtU. Thanks Dr Kitty

Such great content!

Thank you very much for all your hard work. Your channel is a goldmine. 👑

pls bring a 'vision transformer architecture explained' video

Thank you for your selfless sharing and hard work! This video helps me a lot!

Thanks for the amazing video. Can you make video explaining and proving the universal approximation theorem?

This is gold!

This absolutely is gold. Actually it's closer to diamond than gold. My favorite part is 2:44:50, 2:47:50 when the cat also feels the content is wonderful.

Thank you Umar, you video made me connect all the dots!