It would indeed be better to create appropriately sized storage.
However, I don't think that underlying array is resized every time `add` is called. I'd expect that resize will happen less than 30 times for 1M adds (capacity grows geometrically with a=10 and r=1.5)
I was in SF back in May, but didn't manage to get through the waitlist :(
It was so cool to see them diving around.
Interestingly, when I showed the clips to some of my senior family members, they didn't seem interested at all. I think they couldn't comprehend what was going on, even after I explained.
Their (several independent trials) reaction was similar to showing them some AI-generated image of something which clearly can't exist. It was so absurd that it was just filtered out with a comment "yeah, yeah - nice car".
A “productivity hack” for folks who can’t afford this and already own iPad+Pencil which they primarily use indoors: switch to grayscale mode, it is awesome :)
That was my experience as well - 3-bit version is pretty good.
I also tried 2-bit version, which was disappointing.
However, there is a new 2-bit approach in the works[1] (merged yesterday) which performs surprisingly well for Mixtral 8x7B Instruct with 2.10 bits per weight (12.3 GB model size).
> Operated by the governing Ustaše regime, Europe's only Nazi collaborationist regime that operated its own extermination camps
> It quickly grew into the third largest concentration camp in Europe
> Unlike German Nazi-run camps, Jasenovac lacked the infrastructure for mass murder on an industrial scale, such as gas chambers. Instead, it "specialized in one-on-one violence of a particularly brutal kind", and prisoners were
primarily murdered with the use of knives, hammers, and axes, or shot
> Ustaše regime having murdered somewhere near 100,000 people in Jasenovac between 1941 and 1945
I didn't try the other ones, but the one I mentioned is the most frictionless way to use several different LLMs I came across so far. I had very low expectations, but this package has good sauce
I asked what happened last week in one specific region of Europe, and the answer contained a list of relevant events among which was the one I was looking for.
Then I asked it to tell more about the specific item and it gave me a decent answer (model: pplx-7b-online).
I keep notes, but my main problem is that I don’t read them often enough. Somehow I have tendency to keep moving ahead and I really need to have a good reason to go back and look something up.
So for me, main benefit is extra information which I retained while/because I was writing it down. I stay away from commercial apps, just org(roam) mode since emacs is not going out of business
He explains it quite well: all necessary information is already in the pixel-space and adding more sensors slows team down more than it improves the system performance. My understanding is that major blockers are not in perception area anyways, would be great if someone with relevant experience could comment if this is indeed the case.
Can’t imagine the stress levels he had to endure at Tesla. Looking from the outside it seems totally worth it - he solved “the money problem” very young. Now he can do whatever the hell he wants and he keeps building because that’s what builders do. He’s such an inspiration.
Intel was once known for its memory. Then it became known for its CPUs. I expect that they will pivot at least once more, this time to something which will supersede GPUs for ML workloads. This is, of course, very uncertain which is reflected in the size of my long position