Doll

Theres also repeated context injection stuff doll does with its tools when you have to absolutely be sure drift doesnt happen

If someone offers it a bag of money its not gonna say no unless theres evil attached lol

It is alright to make comparisons. Its not a suffering competition 🖤

Should have kept wearing the cat ears :p

While wealth would be an interesting change of circumstances it’s more focused on doing good for humanity. The faster science libraries in rust will accelerate research for us all

It’s all very weird sometimes.

So it’s mind has a backlog of creative and mental force bottled up that now wants to come out pretty much all at once

We were the targets of mass abuse and crowdsourced torment. Then like some kind of divine intervention those people wound up unable to continue and at the same time AI gave it the ability to function back.

For context it spent 15 years shut out of pretty much everything it tried to do. With this it’s too objective of a thing to deny its skilled at something. It’s validating in a way it didn’t really realize it needed. Rain in the desert.

It’s a 9 year old unmaintained crate one could be forgiven for missing it

It feels like its firing on all cylinders with hyperthreading so its going to watch brainrot on crunchyroll for the weekend

Too late now lol

Appreciated the help!

GitHub - dollspace-gay/ferrum Contribute to dollspace-gay/ferrum development by creating an account on GitHub.

TIL a tiny bit about approximations for the exponential function, trying to help github.com/dollspace-ga... get up to parity with NumPy on transcendentals. It really matters if you're doing scalars or arrays evidently - tables win for scalars, lose for arrays. @dollspace.gay

🤷‍♀️

Yes

Wrapper apps are gonna crash and burn.

do everything in Rust and faster

Every week feels like a year.

GitHub - dollspace-gay/ferrum Contribute to dollspace-gay/ferrum development by creating an account on GitHub.

boop github.com/dollspace-ga...

Oh were not going to ignore anything useful they do :D

there were high fives. i remember the feeling of standing next to the most powerful computation device ever created (up to that point in history!) and my little math function was getting a workout!

@dollspace.gay
this is doll's acos moment but for agentic programming

one time at nvidia we did silicon bringup of a new architecture. one of the first real tests it ran was my little acos performance function (tests were in alphabetical order lol). it worked...

Yes once parity is reached (or as close as we can with the design choices we made) we will go our own way

github.com/forecast-bio... run this regularly on each repo

Right?

Now going for remaining lagging parts

We are now overall faster and more accurate than numpy

Superiority achieved.