eoerl

8 ms·

1.
by eoerl 2mo ago
this is amazing https://godbolt.org/ , you can just paste a function or a bunch of them and instantly see what is generated. It doesn´t take being an expert to start to recognize the (auto) vectorized bits
2.
by eoerl 1y ago
Optimization is post hoc here : you have to train first to be able to huffman en ode, so it's not a pure format question
3.
by eoerl 2y ago
In a lot of countries there are rules, for instance limitations in terms of spending or similar time on air for all candidates. I don't know whether that's the case in Romania, but it is completely possible to rule an election out
4.
by eoerl 3y ago
These can also be used for machine learning actually (see Dali for data loading for instance)
5.
by eoerl 3y ago
We (The Eye Tribe folks) sold one at 99$ years ago. 1k-3k is mostly lack of competition I believe.
6.
by eoerl 3y ago
There are flags[1] for that indeed. It feels like half of the people commenting here don't know all that much about the topic they're commenting upon 1: https://pytorch.org/docs/stable/generated/torc
7.
by eoerl 3y ago
> It's not perfect but a group of aligned people in the same physical working space will just dominate a similar group spread apart that has to use chats & zoom to communicate. Management has got to be seeing this, in various fo
8.
by eoerl 4y ago
It is not typically possible to blend models like that, since the training process is (lateral) order insensitive, as far as the model goes.
9.
by eoerl 4y ago
identical outputs, up to float computation shenanigans (not computed in the same order, strictly speaking)
10.
by eoerl 4y ago
yep, same approach but it arrived 3 days later and there's no mention of the [original PR]( https://github.com/huggingface/diffusers/pull/532#issuecomme... ), nice. Else the kernels used in that case -upst
11.
by eoerl 4y ago
did you even peek at the link ? There's a PR on diffusers, and it's mentioned on the front page https://github.com/huggingface/diffusers/pull/532#issuecomme...
12.
by eoerl 4y ago
That's the explanation behind https://news.ycombinator.com/item?id=32985716 , nice work
13.
by eoerl 4y ago
I'm utterly fascinated, big knot fan and I didn't know this one, but it's a combination of parts I knew. It's really strange to me though, and it's probably a culture thing, because for this use case literally every
14.
by eoerl 4y ago
they don't need to layoff, just freeze the hires and wait for the churn
15.
by eoerl 4y ago
actually it seems that it was because a lot of other well known models are not yet supported, missing ops in the Metal backend
16.
by eoerl 4y ago
I'm not sure what you meant with the link, but the parent is right, so adding an explanation here: M1 Ultra has about 400GB/s theoretical bandwidth but Anandtech shows that none of the SoC blocks can actually reach that, pretty fa
17.
by eoerl 4y ago
How much of that have you actually benchmarked, time wise ? I think it’s the point of the article, most of our “benchmarking” is actually marketing. To take an example I’ve been using a cast iron pan for years after having used Teflon all m
18.
by eoerl 4y ago
I’m using a similar setup (G15) for “AI” dev on Linux, it works very well with the Asus-laptop utils. To complement other points: - you can put the nvidia card in “compute” or “hybrid” mode, which removes the need for X restarts. Compute is
19.
by eoerl 5y ago
pushing all this to "un-optimization tax" is an easy pass on apple. - nvidia really is a software company, it's the running joke in the industry. when you buy a nvidia gpu, you pay for the drivers & the frameworks (cuda,
20.
by eoerl 7y ago
Amazon owns nothing, not the ARM IP nor the manufacturing chain (TSMC or Samsung most probably), in that field it's not a big player. It owns what it does with Graviton, that's pretty much it. Else yield obviously counts, that&#x2
21.
by eoerl 7y ago
the key issue when compared to Epyc is that this is mono-die, and not much faster (even with metrics straight from Ampere). Mono-die means that the die is huge, the yield is low, it's probably pretty expensive to produce (and the reaso