gpjt

Writing an LLM from scratch, part 16 – layer normalisation

1 point by gpjt 1 day ago | 0 comments

Leaving PythonAnywhere

3 points by gpjt 1 month ago | 0 comments

Writing an LLM from scratch, part 15 – from context vectors to logits

7 points by gpjt 1 month ago | 0 comments

Writing an LLM from scratch, part 14 – the complexity of self-attention at scale

1 point by gpjt 1 month ago | 0 comments

Writing an LLM from scratch, part 13 – attention heads are dumb

351 points by gpjt 2 months ago | 67 comments

Writing an LLM from scratch, part 12 – multi-head attention

3 points by gpjt 2 months ago | 0 comments

Writing an LLM from scratch, part 11 – batches

2 points by gpjt 2 months ago | 0 comments

The Business of the AI Labs

19 points by gpjt 2 months ago | 3 comments

Writing an LLM from scratch, part 10 – dropout

90 points by gpjt 3 months ago | 8 comments

Adding /Llms.txt

1 point by gpjt 3 months ago | 0 comments

Writing an LLM from scratch, part 9 – causal attention

4 points by gpjt 4 months ago | 0 comments