Back to Basics

Stanford CS336: language modeling from scratch

i've spent enough time using ai, learning some basic stuff. now i wanna understand what the hell is actually going on underneath.

so i'm going through stanford cs336 - language modeling from scratch.

there's a friend of mine who recommended this course to me (@sohxmmmm). he knows every stanford course code by heart, but doesnt know shit about the courses we've got in our university.

the playlist

this is the roadmap we're following:

what i'm trying to figure out

not just how to use a language model.

i wanna understand the actual stuff behind it -> tokenization, data, transformers, training, optimization, evaluation and all the weird little details hidden behind api.call().