Fushuan [he/him] @ fushuan @lemm.ee

Huh?

---

Read more

Posts

2

Comments

1,238

Joined

2 yr. ago

I think that it's more about the fact that the more chrome you add, the less human you are, and that should have a very heavy impact in your mental health, just like it has on people with several implanta nowadays.

It's the premise, to capture them alive so they can undergo a revolutionary treatment aka therapy.

Each and every one of the cases are cases where people had psychotic breaks that can be understood without all the techno magic.

In fact, that's more of an indication that all the cybernetics don't really affect the brain more than whatever mental issues people that get implants would have. If mental issues can't be fixed with a chip, it's very cyberpunk-esque to discard the person ASAP.

Once the file is big enough wouldn't it be better to convert it to json before editing, then converting it back?

Let the computer deal with indents and all that stuff.

There's no better big O sorting method for generic lists. Heap sort has better averages but the big O is the same in the end.

I tried it with protonge 9.something, their current latest, and it worked pretty well. 3080 so woth access tp dlss I had to change it to performance but with it the benchmark finished with excellent too.

Anyhow, saying that frames were a bit worse with a resolution higher than 1080p and frames higher than 60 makes me want to shot on your elitism a bit but I won't. <3

If reactionary stuff is the latest rage and all the progeessives are leaving the site, it's no wonder that the reddit demographic is turning conservative.

Their github releases have the apk available so you can manually download it and install it or use obtainium.

https://github.com/signalapp/Signal-Android/releases

I would glare you back and explain for the 98th time that in my field software being okay doesn't mean that it will run. Data quality and config files that are not out purview can make correct software crash.

It's not data engineering in this case though so I mostly agree with you.

No source but I remember musk saying that the tariff change would require a rough transition that Americans would need to go through to get to good times or whatever.

I though that that was to prepare their base to the actual shitstorm and to them promise them the good times or whatever.

Is this PR comment forever gone now that they are in power?

How is it that there's a whole class country with less native speakers than here in the Basque Contry, a province of Spain?

Maybe that's why bro brought girls.

I don't want to generalise but... Yeah... That's a big reason why a lot of manhuas just suck.

Thanks for the recommendation! I liked Feng Shen Ji too to be honest. It's one of my favorites haha.

It's just... So deserved, you know? Sometimes you can't but laugh in the face of such karma and fucking irony.

Not cuda, but a lower level nvidia proprietary API, your point still stands though.

They replaced it with a lower level nvidia exclusive proprietary API though.

People are really misunderstanding what has happened.

That's the problem, I love wuxia, I love stupid power levels and all that, I love my trash comics. It's the story they tell alongside that that irks me to no end.

Close to zero decent dialogue between people besides shitty villains talking big until the hero flips his eyelash and suddenly everyone is bowing to them. There's no interesting characters, no banter.

To give some examples of wuxia done right (I like the top batch more but the other ones are alright):

Return of the mount hua sect Infinite levelling: murim Return of the crazy demon Nano machine Absolute sword sense Star instructor or something

I reincarnated as the crazed heir Path of the shaman

Half wuxia: Murim login

All of these have a mc with personality, there's decent banter between characters, there's an interesting plot, plot changes make sense...

There's also the "dungeon" theme with portals that grew in popularity since solo levelling (around then) that is basically the same trash plot with another power level vector.

It do be like that with manga/manhua tbh. I still prefer Korean webcomics because I like my trash but... It's been ages since I read a decent Chinese webcomic or manhua. They are so fucking trash. I'm sure there are some good ones but... Yeah.

Not enough for it to make results diverge. Randomness is added to avoid falling into local maximas in optimization. You should still end in the same global maxima. Models usualy run until their optimization converges.

As stated, if the randomness is big enough that multiple reruns end up with different weights aka optimized for different maximas, the randomization is trash. Anything worth their salt won't have randomization big enough.

So, going back to my initial point, we need the training data to validate the weights. There are ways to check the performance of a model (quite literally, the same algorithm that is used to evaluate weights in training is them used to evaluate the trained weights post training) the performance should be identical up to a very small rounding error if a rerun with the same data and parameters is used.

Holy shit thanks I wasn't getting it.