harshuljain13

3 ms·

1. 1.▲
   Free LLM inference handbook: 100 engineers cloned it in week 1(github.com)

   2 points by harshuljain13 4mo ago|0 comments
2. 2.▲
   Why LLM decode is memory-bound, not compute-bound(github.com)

   5 points by harshuljain13 4mo ago|0 comments