GPT-1 Paper Explained

❤️ Support the channel ❤️

Paid Courses I recommend for learning (affiliate links, no extra cost for you):

✨ Free Resources that are great:

💻 My Deep Learning Setup and Recording Setup:

GitHub Repository:

✅ One-Time Donations:

▶️ You Can Connect with me on:

#deeplearning #deeplearning PaperReview #PaperExplained

[AladdinPersson]

I saw some people wanted from scratch implementations & particularly on previous BERT video. Implementing these (correctly) takes a lot of time so it's not always worth the time investment so we'll see if I'll do it.. Right now I'm focusing more on Kaggle and will have more videos on that soon :)

[Frnk]

I remember watching this video around when it came out. It's insane to see how far we've come in language models

[jeffw]

Great video! If you're able to do the GPT-2 and GPT-3 papers as well that would be great, but this is a solid start to understanding how GPT differs from transformers.

I tend to bounce off of the dry academic papers; sometimes it feels like some of them don't *want* to be understood. But your explainer videos have really helped me get a grip on the concepts so I can go back to the paper and get more out of the details.

[tejasvix]

I like your paper explanation videos very much, they helped me a lot in reading paper as i am a newbie, Would love to see more videos, Thanx a lot

[halilhelvaci]

Excellent video! Hope to see it for the other language models as well

[rog]

So are you planning to code GPT from scratch ?

[stephennfernandes]

Hey BERT and GPT1 use sentencepice representation of sentences while training, but how do they fine tune and work with POS tagging task as they require to classify a word based on the enter word per POS token ?? any idea how its done ???

[sguqnjn]

Great explanation! I got a dumb question regarding section 3.1 equation 2 in their paper. It says "h_l = transformer_block(h_{l-1} \forall i \in [1, n])". Is it supposed to be "l" instead of "i"?

[shambhaviaggarwal]

Hey! Can you make a video on pytorch-lightning? I think it is super useful.

[abhishek_maity]

So are you going to put more video on solving the Kaggle problem Use cases??(ML or NLP using DL) ?