Roark66

49 ms·

541.
by Roark66 3y ago
How is it possible for a company to use a single letter as a trademark? This is completely stupid. Not only it is confusing (when I read this title I thought its about Xorg/X11) I can't imagine them being able to enforce such trad
542.
by Roark66 3y ago
>"It can’t be that an Ed Sheeran stream is worth exactly the same as a stream of rain falling on the roof," Warner Music Group CEO, Robert Kyncl Why not? You've already made millions in profit on Ed Sheeran's content.
543.
by Roark66 3y ago
>But most of what I like about it is great when “visiting,” not living. When I’ve lived in cities that had great public transportation it was occasionally nice; I could go downtown and get drunk without worrying about/coordinating t
544.
by Roark66 3y ago
I'm not sure why do we need one. Nano works fine everywhere.
545.
by Roark66 3y ago
Wait, I wasn't aware llama.cpp even runs on x86_64.i thought it is arm hw only. If what you say is correct that indeed is very interesting. Especially if I can extend it to other models like falcon.
546.
by Roark66 3y ago
I wonder how does it compare to running same model on the same cpu in ram(making sure fast cpu ML library that utilises AVX2 is used for example Intel MKL)? Also, when doing a test like this it's important to compare same bit depths so
547.
by Roark66 3y ago
Not only in desperate places. I heard last year (or the year before) someone stole few km of train wire in Germany. Although to this day some people think it was a Russian sabotage rather than genuine theft. Previously (for example in Pola
548.
by Roark66 3y ago
Cuda runs on a lower layer than torch. There is not much competition to cuda if one wants performance, but if one just wants to run it there is cpu (openblas etc and Intel's MKL) and onnx. It's pretty astonishing to me AMD has nei
549.
by Roark66 3y ago
It is useful to mention running inference on modern cpus that have AVX2 is not that bad. Sure it is slower than on the gpu, but you get the benefit of having a single long continuous region of ram. But there is one huge problem why this is
550.
by Roark66 3y ago
It is an interesting paper. Any idea when the code/data will be released? It appears it has been almost 2 months since the paper was submitted, but the link given leads to a random bing page :-(
551.
by Roark66 3y ago
>It seems like you're saying that people are mainly using adblockers because the recommendations aren't good enough, and if the recommendations were improved they'd respond by turning off the adblock? I can see how what I
552.
by Roark66 3y ago
Private and secure? I thought the main issue with privacy and security of (not at all)OpenAI models is that by using their products you agree for them to retain all the data you send and receive from the models forever for whatever they cho
553.
by Roark66 3y ago
I wonder how are they "detecting adblock". Let's say I have an alternative youtube client like smartyoutubetv. What will they do? Check if ads are loaded? Fine, the software can be written to load the ads, but not display the
554.
by Roark66 3y ago
To some extent I agree, but it really depends on what you do and for whom. When I asked why this company decided to swap windows for linux (for certain teams) I was told their tools (terraform, python, ssh via cygwin, helm, kubectl etc) wou
555.
by Roark66 3y ago
The main thing every LLM should know is that ARM will eat x86_64's lunch in ML. Why? Because of the shared/unified memory model. M2 Ultra from apple can use up to 192GB of RAM. Even your smartphone thanks to this model can run net
556.
by Roark66 3y ago
Thank you for explaining it. I was aware of the problem of too short context length. I'd love to be able to pass an entire programming project with my prompt for example (or a book in your example). I think I understand how it works no
557.
by Roark66 3y ago
Me too. You know what's another good sign times are changing? A fortune 500 company gave me a Linux laptop to WFH. Even a decade ago this would be extremely improbable.
558.
by Roark66 3y ago
Is there a better article explaining how vector DBs are used with LLMs exactly? This 4 point description raises more questions than it answers for me. For example what is the difference between a "chunk vector" and a "prompt
559.
by Roark66 3y ago
>new record for discharge time on Wendelstein 7-X: the hot plasma could be maintained for eight minutes. This is of course a milestone, but note they're currently testing the heating system. It doesn't mean there was any fusion
560.
by Roark66 3y ago
I suppose for every situation where a pilot gets confused or panics that results in a serious incident there are dozens(hundreds?) situations that never progressed far enough, because a pilot calmly prevented it. Personally I consider autop
561.
by Roark66 3y ago
In this I'm mostly concerned by the extremely vague description of what is it you can't do with this new terraform commercially. You can't "compete" with hashicorp. Well, if I'm doing devops consulting and I wr
562.
by Roark66 3y ago
That's interesting. I know not everyone is in position to pick and choose between jobs, but frequently you can improve your "it situation" a lot if you raise it during recruitment. I recently started working for a (large) com
563.
by Roark66 3y ago
>Not a hoarder but I do feel a little twinge in my heart if I see someone throw away plush toys or other things with faces. :-) Interesting, I get the same, but for electronics/anything mechanical. Also the longer I have some device
564.
by Roark66 3y ago
It's hilarious on Linux you need chrome to use teams for screen sharing / video to work, and Firefox for sharepoint file sharing to properly work. If someone shares a file on a channel, you can open it in chrome. But you can'
565.
by Roark66 3y ago
Why is it that most "chat/conference" apps become horrible sooner or later? I still remember when teams used to be an OK app. It even had a Linux desktop client. I remember when slack was actually fast, I remember Skype out b
566.
by Roark66 3y ago
So, let me get this right. CNET started using "generative AI" to write their articles. Google no doubt detected it and down ranked them to hell. CNET stopped the AI generation and they decided to delete their archives to improve t
567.
by Roark66 3y ago
>I'd do it because I think it's better to have separate work / life areas; right now my WFH spot is also my PC gaming spot and the bedroom and the place where laundry dries, and I wouldn't mind more distance between w
568.
by Roark66 3y ago
>My vaguely populist take is that the pandemic, WFH/remote and hybrid gave the mass white collar class a taste of what the C-suite has had for years.. flexibility. Exactly. I've been working 100% from home since 2017 (3 years b
569.
by Roark66 3y ago
I guess it depends on region. I've been using Google meet, zoom and teams and by a large margin teams provides the best conference experience(best audio quality, less interruptions etc). Shitty web app and lack of desktop app on Linux
570.
by Roark66 3y ago
Pro life tip: all medical doctors are like that. When your health is on the line question everything. Yes, doctors are just technicians trying to fix your body. In your It career how many lousy technicians have you met? This is exactly the

More ›