Got my hands on a Jetson Orin Nano, thanks to a teacher who pushed me to explore a bit more. Setting it up now. Curious to see what comes out of this.

@stanfordnlp @tatsu_hashimoto Papers suddenly feel less scary once the architecture clicks:)

Finished lecture 3/17 of CS336. This one really did it. Suddenly I want to read all the LLM papers I’ve been seeing referenced everywhere. Probably a terrible idea. Probably going to do it anyway. (notes are getting messier, understanding is getting better)

Spent time studying attention, self-attention, and transformers through CS224N (lectures by @chrmanning & @johnhewtt). Conceptually it makes sense now. but every time matrices show up my brain goes: “yeah okay, we should revise this again” 😅 Understanding ≠ mastery.

@percyliang @tatsu_hashimoto ohhh

Thought I’d just “watch a lecture” and understand LLMs. Nope. 1h20m video → ~4–5 hours of pausing, rewinding, and filling pages like this. 2/17 CS336 done. Gonna finish CS224N lectures 6,7 & 8 before jumping into LLM architecture + hyperparams. Kinda slow. Enjoying it.