MrBra

What's the genie effect?

For me it's been super stable. I've hardly seen any bugs. And in those remote cases, it would be more correct to call them quirks than bugs, which have later been fixed anyway. I've been using for intensive gaming, AI projects, and audio production. And when I say audio production I don't say Audacity. I say recent versions of Ableton Live running on ASIO drivers with windows VSTs and Max 4 Live instruments at 5 ms latency, all of this running through Wine with an amazing Wine managing software called Bottle (hehe). As for gaming,, it's not hard to see people claming they get even more fps than they get with windows. It's not a PopOS thing, it's the Linux ecosystem that is finally getting mature enough to pull this out (this time for real). On top of this, System74, the company behind PopOS who is selling laptops with that OS, are also optimizing the kernel to make sure everything runs super smoothly... I really don't see where your "buggy as hell" is coming from.

we don't need an FOSS self-hosted alternative to X. We need a FOSS alternative to the cloud/hosted model such that there's distinction between the two, everything is everywhere for everyone, free. Fediverse, but better. But we're not even at Fediverse ;)

After switching to OnlyOffice I never looked back.

Also, take a look at Heroic (game launcher and library with integrated Wine for GOG/Epic/Amazon Prime), available for Windows too.

Improvements in inference speed would also manifest itself on those bigger models that may only partially fit into GPU VRAM. In some cases, the improvement on the GPU side alone, is strong enough to basically turn what you would previously consider a too-slow-to-be-usable higher quality model, into a faster, usable one.

What's a peleton? Genuine question from a non native English speaker.

Also you can right-click a playlist and select "Exclude this from musical profiling" (translated from my native language). I did that for many older playlists I'm no longer into, and this seemed to cause a significant steer in weekly recommendations which have since turned to be much more in line with my current tastes.

But the Clone-a-Lisa page does not use a native dark theme, and the problem seem to be just with the paint colors, which in my case for some reason are altered by Dark Reader. It could also be that it was a temporarily glitch as I've noticed that color manipulation by Dark Reader can be inconsistent at times (could be due to DOM loading timing and stuff like that).

This doesn't work well with browser extensions which darken web pages color themes (like "Dark Reader"), as all the painting colors will be rendered to black.

I know It might be obvious but having those extension always on I tend to forget they sometime can cause some visibility issues, so I'm leaving this here :)

Or better: the web ate the OS resources. Or better: an app ate the OS resources.

Which is exactly what resources are for, when eating is performed correctly :)

Ok, now take the number of years during which Firefox was slower and calculate an averaged value of its speed. This will account for all the frustration it would have caused switching to it in the past. Being faster for 2 days, 2 weeks, 2 months, and probably even for 2 years wouldn't be enough to cause a big change in that averaged value. And that would also measure how much of a guarateed good choice would be switching to Firefox right now. As soon as it will stay faster for enough time to smooth out the differences then I'll start consider using it.

I've been a Firefox advocate and user for many years, but then about 6 years ago I had to face the hard truth of how bad it performed with lots of tabs open, and how much of a pain I was causing myself for not switching to a Chromium powered browser.

About one month ago, I've checked it out again under the same conditions... and sadly it's still not even close to competitors' memory management.

What's the best LLM I could run on a 12 GB VRAM GPU (Nvidia) ?

The pandemic made it so much worse for a lot of people I know. Or should I say it was exactly those people who made it worse for all of us? I am sure that a certain slice of the population my age

Where's the demo number? Can't seem to find it?

Does anyone here use the free Feedbro extension? I'm asking because it's working really good for me and I've never seen it mentioned here when it comes to RSS feed clients.

Is it possible to setup PayPal to automatically forward received money to a bank account or a credit card, without allowing them any withdrawal rights on that bank account or credit card?

Would it be safe to use it that way?

Any idea on how long could it take to get this in the stable channel?

If you are only interested in Stable Diffusion there's no need to get up to 24 GB. Plenty of people are using a 12 GB RTX 3060 (with that you could generate something like 4 simultaneous 512x512 images in a matter of seconds).

Even the cheaper and recently re-released 12 GB RTX 2060 could be a good pick although the 2060 is generation behind and is less efficient in terms of electricity consumption.

Of course with more GB you get higher resolution images, but plenty of people just generate at 512x512 and then use other AI projects (for example "real-ESRGAN") to later upscale their images, which will still let you achieve great results.

I think a good advice could be to join Stable Diffusion Discord and talk to other users sharing their results and experiences there.

By the way, IIRC (please double-check the following) it could be also worth noting that the guys behind Stable Diffusion (Stability.AI) declared that in the end they will eventually bring down the VRAM usage to approximately 5 GB.

However more GB are always a good thing in general for Machine Learning...

My bad! Yes Waifu2x is just a single algorithm, you are right.

The confusion originates from the fact that I was using a GUI project for Waifu2x called "Waifu2x Extension GUI" (https://github.com/AaronFeng753/Waifu2x-Extension-GUI) which other than Waifu2x also supports other algorithms like Real-ESRGAN, Real-CUGAN, SRMD, RealSR, Anime4K, RIFE, IFRNet, CAIN, DAIN, and ACNet.

So as you said Cupscale is surely more advanced than Waifu2x (the single algorithm), but do you think it's also better than Waifu2x Extension GUI?

As I said Waifu2x too supports many different models aimed at different content, so what's the big improvement with Cupscale?

I can run DALL-E mini with the MEGA model with 12 GB of VRAM. What are the requirements for minDALL-E and MEGA, in terms of VRAM?

How is minDALL-E better than DALL-E mini?

I can run DALL-E mini with the MEGA model with 12 GB of VRAM. What are the requirements for minDALL-E in terms of VRAM?

How much VRAM will minDALL-E plus the MEGA model take?

Thanks to the amazing @lucidrains there's already an open-source implementation of DALL-E 2: https://github.com/lucidrains/DALLE2-pytorch and a pretrained model for it should be released within this year.

The same person is also at work on an open-source implementation of Google's Imagen which should be even better (and faster) than DALLE-2: https://github.com/lucidrains/imagen-pytorch.

This is possible because the original research papers behind DALLE-2 and Imagen were both publicly released.

I was reading on Reddit that you would only need 12 GB of VRAM for DALL-E 2 to run on a local machine https://www.reddit.com/r/dalle2/comments/w3sbt3/comment/ih01....

I've recently updated waifu2x and I've seen it now supports lots of algorithms for different use cases and contexts and it also supports other tasks like frame interpolation. So could you briefly explain in what is cupscale better than it?

I think he refers to the premium version of Waifu2x.

I just tried it and I'm not impressed at all. There are many better, yet still free alternatives, so not sure I understand why this got on front page...

This is even better though: https://www.99xtechnology.com/blog/research/neutralinojs-int...