📰 Hacker News · gpjt

LLM from scratch, part 28 – training a base model from scratch on an RTX 3090. 121 comments, 540 points on Hacker News.

Writing an LLM from scratch, part 22 – training our LLM. 10 comments, 254 points on Hacker News.

The maths you need to start understanding LLMs. 120 comments, 616 points on Hacker News.

Writing an LLM from scratch, part 13 – attention heads are dumb. 67 comments, 351 points on Hacker News.

Writing an LLM from scratch, part 8 – trainable self-attention. 31 comments, 380 points on Hacker News.