Vicki

vickiboykis.com

17 December 2024 20:52

New post: I think the debate around whether LLMs can replace developers is lacking some nuance because, if we're lucky, we as developers don't work without context. I work with my Alphabet Radio on, streaming the greatest hits of the best devs I've worked with. vickiboykis.com/2024/12/16/w...

Ethan

ethanrosenthal.com

17 December 2024 03:42

This is my preferred way of doing things right now. The CLI only selects the config. All of your config params are grouped together in a single object. You get nice type hints and validation from pydantic. Your script contains a record of previously-run configs.

Christian Minich

christianminich.com

17 December 2024 02:24

When choosing between parallelizing within your orchestrator or out to a parallelization engine like dash, “choose a breakpoint that makes sense to a human operator” is great guidance. Applies to lots of aspects of designing systems. Human operators should not be an afterthought!

Niko Föhr

fohrloop.com

15 December 2024 05:05

Monte Carlo methods use random sampling to solve tough problems (optimization, integration, ...) @13:55 When your simulations are expensive, you want to be clever about sampling (Quasi Monte Carlo) -> less samples, same info @32:00 cleverness of different sampling methods compared #datascience

Kapalı içerik