Loading video...

Video Failed to Load

Go Home

llm.c by Hand✍️ C programming + matrix multiplication by hand This combination is perhaps as low as we can get to explain how the Transformer works. Special thanks to Andrej Karpathy for encouraging early feedback and tetsuo //: 👾 for helping me understand the pragma magic. I hope this...

302,625 views • 2 years ago •via X (Twitter)

11 Comments

Yuchen Jin's profile picture
Yuchen Jin2 years ago

@karpathy @7etsuo Awesome! Looking forward to an llm.c Attention episode

Tom Yeh's profile picture
Tom Yeh2 years ago

@karpathy @7etsuo This is a great suggestion! Thanks!

Antonio Linares's profile picture
Antonio Linares2 years ago

@karpathy @7etsuo Is there a way to load a single layer of a LLM (in example Llama3-70B) and process it layer by layer so it fits on a normal PC memory ? Could you explain how to do it ?

Tom Yeh's profile picture
Tom Yeh2 years ago

@karpathy @7etsuo It is possible. But it would be quite slow. To generate each token, the model has to evaluate every one of the hundreds of layers. Say loading each layer takes 1 second. It would take 100 seconds to load 100 layers for one token. Then you multiple that by 1000 tokens. Very slow.

rakesh's profile picture
rakesh2 years ago

@karpathy @7etsuo Goated acc

ed's profile picture
ed2 years ago

@karpathy @7etsuo Amazing 🔥

Cuhmunity Notes's profile picture
Cuhmunity Notes2 years ago

@karpathy @7etsuo 🔥

Byron Hsu's profile picture
Byron Hsu2 years ago

@karpathy @7etsuo This is awesome

Filippo Broggini's profile picture
Filippo Broggini2 years ago

@karpathy @7etsuo Is this also available elsewhere online?

George D Gregory's profile picture
George D Gregory2 years ago

@karpathy @7etsuo Can these be made into the notebooks!? That would be awesome.

Tom Yeh's profile picture
Tom Yeh2 years ago

@karpathy @7etsuo I am waiting for a notebook ninja to step up to help.😄

Related Videos