Re:Reproduce GPT-2 (124M) in llm.c in 90 minutes for $20
by karpathy on Tuesday May 28, 2024 @01:56PM
(#40503512)
Yes definitely. Related tweet of mine:
https://x.com/karpathy/status/1760388761349927356?lang=en
1. Build the thing
2. Build the ramp
Currently on step 1 :). It helps to build it first so you know where you are going, and then you can more easily re-build it when you're vector pointed at the end result.
Re:Reproduce GPT-2 (124M) in llm.c in 90 minutes for $20
by htrp on Tuesday May 28, 2024 @02:32PM
(#40503933)
Everytime you take gardening leave, you build something new and interesting!
Re:Reproduce GPT-2 (124M) in llm.c in 90 minutes for $20
by lagrange77 on Tuesday May 28, 2024 @02:15PM
(#40503739)
That's fantastic. My gradient field is pointing towards it.
Thank you again!
Re:Reproduce GPT-2 (124M) in llm.c in 90 minutes for $20
by LorenzoGood on Tuesday May 28, 2024 @11:03PM
(#40508012)
I love when you leave your job.
> Ultimately my interest in llm.c is to have a nice, clean, minimal, super dependency-light repo in direct C/CUDA implementation, which I find aesthetically pleasing.
Also, it's awesome that you spend your time on your passion.
Any plans on making a video series on llm.c? :D