renonce

22 ms·

241.
by renonce 4y ago
They are actually worth that amount if they hold them though. They are worth one 2033 dollar or 0.7 2023 dollar. 2033 dollar and 2023 dollar should be considered two different currencies. The bonds are pegged to 2033 dollars but customers
242.
by renonce 4y ago
> The original idea with a 100% collaterilised peg was that the entire reserve would be cash. What is cash? Banknotes? You can't store 500 million pieces of $100 banknotes easily or safely. Cash usually refers to deposits at accredi
243.
by renonce 4y ago
It's officially banned but it's not like they have been banned successfully. You just buy crypto by transfering money P2P to another individual and that individual sends you crypto back. The government has no way of knowing that a
244.
by renonce 4y ago
Weights are more valuable than training code in one regard. Even with the training code you may not have the dataset and reproduction requires a massive GPU cluster that few can afford.
245.
by renonce 4y ago
Anyone can submit a request form and enter their email address to request access to the weights. People who have a .edu email and are involved in deep learning are likely to succeed and get a download link sent to their email.
246.
by renonce 4y ago
It's fairly easy to obtain the weights. I've had two of my friends downloading these weights and sharing them with me, so it's probably not surprising that the weights got leaked.
247.
by renonce 4y ago
You do need to guess things like activation functions, number of attention heads, order of attention layers, etc. Often the parameter names reveal something about these.
248.
by renonce 4y ago
There is a lot of manual process involved such as writing training scripts, scarping and processing training data, choosing the best weights among several runs, and spending lots of costly computation. Maybe these should make it copyrightab
249.
by renonce 4y ago
I wonder if there is a way to simulate a balloon in reverse, i.e. given its destination figure out which sources it could come from. Maybe it would just be a matter of reversing the winds and time? Would be interesting if it were doable.
250.
by renonce 4y ago
How long does it take for you to download 80TB? From what I can see Google allows you to download 10TB per day but who knows when they will change that limit.
251.
by renonce 4y ago
Mandarin speaker here. There are indeed different counting words for different nouns, so for a bird (鸟) you say 一只鸟 and for a chair (椅子) you say 一把椅子.
252.
by renonce 4y ago
The owner of a simple hobby project like this, without SEO or sign-in or whatsoever, probably don’t need to care so much to protect his domain. If StackOverflow asks for his domain he can just give it away and set up the site at a new domai
253.
by renonce 4y ago
Microsoft is well known to be organized as a set of teams that hate each other. Azure and Outlook are likely two different teams. There is a famous comic for that: https://i.imgur.com/XLuaF0h.png
254.
by renonce 4y ago
Check out Syncthing. I personally run two instances of syncthing, synchronization happens very fast and as long as there isn't high write volumes it syncs pretty quickly. If you don't need realtime backup, rsync'ing in a loop
255.
by renonce 4y ago
> when my laptop takes 10 minutes to start moving the mouse cursor smoothly after waking from a weekend suspend Apologies for not reading your comment fully, but in this specific instance it's because a weekend suspend causes hibern
256.
by renonce 4y ago
Without retraining the model, you can often get a different image by generating it again or modifying the prompt slightly. Makes people wonder if changing the training set would have any bigger impact.
257.
by renonce 4y ago
The observable universe has only 10^80 atoms. A small image of 128x128 pixels has more variations than that.
258.
by renonce 4y ago
That's called inflation.
259.
by renonce 4y ago
Many TLC drives have SLC cache so if you limit usage of the drive you may effectively use it in SLC-only mode. This does not say anything about its endurance however. If you want endurance, better mirror your data to HDD.
260.
by renonce 4y ago
Bitcoin price is at $23,530 right now. Imagine Alice wants to buy 1 bitcoin and sell it a few minutes later for some profit. In that case Alice can apply a 10x leverage since it is short term, and needs only $2,400 as collateral. 10 minutes
261.
by renonce 4y ago
https://aviationsourcenews.com/breaking/atlas-air-to-fly-spe... The ceremony is prepared by CEO himself in this case.
262.
by renonce 4y ago
I'm an undergraduate who took a course based on this book. While I agree that the book is not good at being a "foundations" book, I would say the ideas taught in the book are very helpful in building an intuition for data sci
263.
by renonce 4y ago
Did you try that in practice?
264.
by renonce 4y ago
I’ve been using Arch Linux for several years now. It has a rich ecosystem of packages, a wiki for many aspects of the system (you might be using the wiki even if you came from another distribution), and AUR fills the hole for whatever packa
265.
by renonce 4y ago
Are all these 235 microservices necessary for the bulk of features that 99% of users are using? AWS is not the only S3 provider, nor the best, and it would be interesting to see how much engineering is involved in other S3 providers.
266.
by renonce 4y ago
You can have such redundancy for wire too. People simply figured out that it’s so reliable there is no need for extra redundancy.
267.
by renonce 4y ago
Maybe too late to give you any helpful advice, but setting up Advanced Protection may make sense. You need to buy at least two (preferably three) YubiKeys and the password plus any of these keys allow you to login to your account. Nothing m
268.
by renonce 4y ago
If the data was generated from a generator of constant complexity, its kolmogorov complexity cannot exceed the complexity of the generator. As the amount of data increases, the kolmogorov complexity will increase until it converges after a
269.
by renonce 4y ago
AWS is the worst offender. They don’t allow you to set hard limits on billing, claiming that is for your own good, while the fact is they are putting all your data on hostage so they could force sell you things at ridiculously high traffic
270.
by renonce 4y ago
I think they are just reusing technologies commonly used in China on TikTok website so they don't write it twice. In China many frontend developers write code for WeChat mini program. WeChat mini program is just a typical HTML website

More ›