acidflask

5 ms·

1.
Languagemodels.jl – Talk to language models programmatically from Julia(github.com)
3 points by acidflask 3y ago|1 comments
2.
by acidflask 3y ago
A port of Andrej Karpathy's llama2.c to pure Julia. Works on Llama2 and Andrej's tinyllamas models. REPL mode allows interaction from the Julia console
3.
A CSS stylesheet for typographically improved Jupyter notebooks(github.com)
1 points by acidflask 6y ago|0 comments
4.
Oil prices drop below $5(finance.yahoo.com)
3 points by acidflask 6y ago|0 comments
5.
Facebook issues correction notice on States Times Review's post(channelnewsasia.com)
2 points by acidflask 7y ago|0 comments
6.
Chika's Test (For Divisibility by 7)(westminsterunder.org.uk)
4 points by acidflask 7y ago|2 comments
7.
Fairness Under Unawareness: Assessing Disparity When Protected Class Is Unobserv(arxiv.org)
1 points by acidflask 8y ago|0 comments
8.
by acidflask 9y ago
How long until we can have Julia wrappers?
9.
Robust benchmarking in noisy environments(arxiv.org)
2 points by acidflask 10y ago|0 comments
10.
by acidflask 10y ago
This question keeps me up at night, as I presume it does for my colleagues at Julia Computing. Not that switching to a fully commercial model is necessarily a bad thing, but Julia Labs, like any academic group, always has to worry about whe
11.
by acidflask 10y ago
See Slide 4 for why it's important to be open source. TL;DR: Researcher A finds things he wants improved in Magma (closed source) but can't. Researcher B tries to write improved FOSS implementation, but lost his job, likely in par
12.
by acidflask 10y ago
I'll be there for moral support, if I'm allowed to crash!
13.
by acidflask 10y ago
And the problem with academic funding is that it's really hard to pay people to improve the documentation with those funds.
14.
by acidflask 10y ago
As Stein explains in the slides: because he couldn't sustainably pay people to work with him on SAGE, and there is only so much you can do alone.
15.
The Origins of SageMath; I am leaving academia to build a company [pdf](wstein.org)
415 points by acidflask 10y ago|196 comments
16.
by acidflask 10y ago
Thanks for the input! Yes, there are at least two separate stages of historical development here. The first is when people realized it was useful to repeat the same operation on different data in memory and viewed the collection of data as
17.
by acidflask 10y ago
Would you happen to have any insight into the question that brought me to go dig through old letters in a library in the first place? https://news.ycombinator.com/item?id=11799143
18.
by acidflask 10y ago
The paper is about, among other things, the history of the array data structure. It's far too early to advertise, but you can see a very early version on my GitHub account. :) I was surprised to discover recently that the word 'ar
19.
by acidflask 10y ago
Thanks for contributing the transcript! I've started a GitHub repo with a first draft of transcripts for all the letters. https://github.com/jiahao/backus-dijkstra-letters-1979 Anyone else who is interested is wel
20.
by acidflask 10y ago
The title "This guy’s arrogance takes your breath away" is taken directly from Backus's own description of this collection of letters. I've changed the title to make clearer that it is a direct quotation. Unfortunately,
21.
Letters between Backus and Dijkstra (1979)(medium.com)
237 points by acidflask 10y ago|122 comments
22.
by acidflask 10y ago
I'm genuinely surprised that one would say that Julia is "slowing down in development". Perhaps it's because less press is being generated about Julia? Or that the commit rate has gone down slightly, now that the easier
23.
by acidflask 10y ago
Making people think is my favorite part of my job. :)
24.
Language complexity, measured by how many rules GitHub's .gitignore has(gist.github.com)
104 points by acidflask 10y ago|45 comments
25.
by acidflask 10y ago
> A big problem is that chained floating point operations are a O(sqrt n) martingale. Mostly false; see https://www.cs.berkeley.edu/~wkahan/Mindless.pdf It is true that an arbitrary, independent sequence of n float
26.
by acidflask 10y ago
Have you actually profiled the Julia code you wrote and ensured that you have followed the performance tips in the manual? If you did it correctly all the time should be spent in linear algebra operations, and so it shouldn't matter wh
27.
by acidflask 10y ago
It would be really interesting if you'd like to implement some quantum chemistry! Rick Muller (of PyQuante fame) has implemented a port to Julia, but I think there's a lot of room to use more language features. https://
28.
by acidflask 10y ago
Actually, chemistry and physics do quite well with nonprofit publishers. Compare the profit margins of for-profit publishers with non-profit ones like the American Chemical Society. In 2015, the ACS made a $10m profit on $500m gross revenue
29.
by acidflask 10y ago
To cite just one example, Elsevier consistently reports over $1b/yr profit on $3b/yr revenue. Not exactly "almost nothing". https://en.wikipedia.org/wiki/Elsevier https://libraries.mit.ed
30.
by acidflask 10y ago
The issues about the lack of tooling and stability are well heeded. Unfortunately, we at the Julia Lab at MIT, like our colleagues that are the JuMP group at the MIT ORC, are limited to working on things that can be justified as research ac

More ›