krapht

8 ms·

61.
by krapht 1y ago
HIP is something you can put in production and Zluda is a proof-of-concept?
62.
by krapht 1y ago
Of libraries that are better than the STL? Or data structures that the STL is missing? Without getting into the weeds of esoteric data structures and algorithms I think there's a few holes that aren't addressed: https:/&#x2F
63.
by krapht 1y ago
So the MI300x has 8 different memory domains, and although you can treat it as one flat memory space, if you want to reach their advertised peak memory bandwidth you have to work with it like an 8-socket board.
64.
by krapht 1y ago
Why should there be HoTT, though? A lot less work has gone into making HoTT theorem provers ergonomic to use. Also, the documentation is a lot more sparse. The benefits of HoTT are also unclear - it seems to only save work when dealing with
65.
by krapht 1y ago
Certainly not for introductory courses. There's already too much on the table - how to prove, how to program, and the base material itself. I believe that's the experience of faculty who've tried as well - it's fine for
66.
by krapht 1y ago
https://docs.python.org/3/tutorial/floatingpoint.html Stop at any finite number of bits, and you get an approximation. On most machines today, floats are approximated using a binary fraction with the numerator usi
67.
by krapht 1y ago
I agree, and yet here we are. If it wasn't built with automake I would even consider getting wheels built on the projects behalf, but I can't make heads or tails of M4. EDIT: But let's not pretend getting cibuildwheel going f
68.
by krapht 1y ago
Conda is still the recommended way to work with Intel's Python distribution, so there's a reason for it to live on my work computer. The fastest graph library I know, graph-tool, needs to be installed with Conda too.
69.
by krapht 1y ago
I tried to give Nix a chance but I never figured out how to package dependencies that didn't already have a derivation written for them. Particularly since I do a lot of ML work - I never figured out how to handle mixed Python/C++
70.
by krapht 1y ago
Because I benchmarked both on my dataset and found that Tesseract was better for my use-case?
71.
by krapht 1y ago
Because the VLM doesn't know it hallucinated. When you get a Tesseract error you can flag the OCR job for manual review.
72.
by krapht 1y ago
I just built a pipeline with tesseract last year. What's better that is open source and runnable locally? VLLM hallucination is a blocker for my use case.
73.
by krapht 1y ago
It is nicer, but that interface latency would turn me off. When I search for online groceries, I want simple queries e.g spinach to return varieties (e.g frozen, fresh, and canned) of spinach as fast as possible.
74.
by krapht 1y ago
I know you're going to reply with "BUT MY PREPROCESSOR", but template specialization is a big win and improvement (see qsort vs std::sort).
75.
by krapht 1y ago
The article title is half-true. It wasn't so much they migrated away from Rust, but that they migrated away from Bevy, which is an alpha quality game engine. I wouldn't have read the article if it'd been labeled that, so kudo
76.
by krapht 1y ago
I'm only familiar with C++, Python, and SQL. Neither GDB nor PDB helps here, and I've never heard of a SQL debugger that will break apart expressions and let you view intermediate query results.
77.
by krapht 1y ago
Yes, but here's my hot take - what if you didn't have to edit the source code to debug it? Instead of chaining method calls you just assign to a temporary variable. Then you can set breakpoints and inspect variable values like you
78.
by krapht 1y ago
Hackers are not crackers.
79.
by krapht 1y ago
Is this similar to what 2, 4 dinitrophenol affects?
80.
by krapht 1y ago
It's so rare to see work at this level posted online. RF experimentation isn't as popular as it used to be. I enjoyed this article!
81.
by krapht 1y ago
Big combinatorial problems still use genetic algorithms. Specifically I know it's still used for logistics routing problems that are too big for solvers like Gurobi. Deep learning on graphs is unfortunately still a little underwhelming
82.
by krapht 2y ago
I agree that the US bears significant responsibility for the ~5% civilian deaths in Iraq. These were through: 1) Direct combat fatalities (~15% of casualties) 2) Failing to stabilize Iraq post-invasion 3) Enabling conditions for prolonged c
83.
by krapht 2y ago
I had thought for awhile to learn Modelica, but the opportunity to use it in a project never came up. More specialized tools were always easier to use - Simulink, or SPICE for electrical circuits.
84.
by krapht 2y ago
That correspondence doesn't automatically mean you get a useful compiler from proofs. Rather, that's the exception instead of the rule.
85.
by krapht 2y ago
I'd like to see proof that Medicare does not evaluate QALY and spends unlimited money in general.
86.
by krapht 2y ago
On that note I've been meaning to explore Wuffs for this problem. https://github.com/google/wuffs
87.
by krapht 2y ago
> Edit: Also, side note, why are so many people running their LLMs in the cloud? All the cutting edge models are open weight licensed, and run locally. You don't need to depend on some corporation that will inevitably rug-pull you.
88.
by krapht 2y ago
It's really easy when you can talk to living people today who experienced the ussr in the 80s and compare to the average Western experience at the same time.
89.
by krapht 2y ago
I feel like I'm not the target audience for this. When I have large data, then I directly write SQL queries and run them against the database. It's impossible to improve performance when you have to go out to the DB anyway; might
90.
by krapht 2y ago
Why do straight to dystopia when notary publics exist?

More ›