Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
marmaduke
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
marmaduke
2y ago
I’m doing some C on esp32 currently, and wondered about switching to rust, and I understand that simple cases with Rust are solid. I’m definitely interested in trying it out. However we need to work with a lot of the hardware apis like twa
32.
▲
by
marmaduke
2y ago
Yep that’s my reading too. I like to see it from a dynamical systems perspective: as a system approaches an attractor, the phase flow slows down, while the wall time marches on steadily. If we consider the “perspective” of the system, wh
33.
▲
Faust, a compiled audio processing language
(faustdoc.grame.fr)
1 points
by
marmaduke
2y ago
|
0 comments
34.
▲
by
marmaduke
2y ago
I like to think of fuse as a way of allowing Makefiles to specify DAGs over arbitrary resources. For instance, a fuse fs exposing the state of a k8s cluster might ease writing operators accessible for simpler minds like mine. Or email, why
35.
▲
by
marmaduke
2y ago
I meant just in terms of compute power. Like my isp gives me a static IP with forward and reverse dns, and the box lets me put the phone WiFi ip address in the DMZ so all traffic is handled by the phone. Then the termux app lets me run ss
36.
▲
by
marmaduke
2y ago
I was poking at this a bit here https://github.com/maedoc/rwkv.js using the Rwkv.cpp compiled with emscripten, but I didn’t quite figure out the tokenizers part (yet, only spent about an hour on it) Nevertheless I am p
37.
▲
by
marmaduke
2y ago
It’s a joke ! You can run an email server off your phone
38.
▲
by
marmaduke
2y ago
Couldn’t a tcc or similarly simple C compiler be used instead of a 100MB Clang? Where’s the C to wasm compiler hiding?
39.
▲
by
marmaduke
2y ago
Very cool. Can you describe any fine tuning or RAG you did? Did you have to go through a bootstrapping process? Totally get it if you don’t want to share details.
40.
▲
by
marmaduke
2y ago
I enjoyed the less serious part a lot. I wish more programming related projects could embrace the whimsical. That might the best way to honor the python tradition in any case :)
41.
▲
by
marmaduke
2y ago
Reading the implementation details kinda got me wanting to do a c11 threads based port.
42.
▲
by
marmaduke
2y ago
That’s what I was wondering. Iiuc openmp uses work stealing and the spice readme talks about the inefficiency of that: https://github.com/judofyr/spice?tab=readme-ov-file#work-ste... so I’d be interested in seeing a b
43.
▲
by
marmaduke
2y ago
Yep and in general I think the matrix got it right when discussing Trinity and Neo's visions, the Oracle states: "We can never see past the choices we don't understand".
44.
▲
by
marmaduke
2y ago
Lots of things I think, lack of understanding, lack of motivation. Typical suspects. I think We look for easy solutions like using LLMs for instance (but not exclusively) doing busywork when we shouldn’t be doing the busywork in the first
45.
▲
by
marmaduke
2y ago
Devils advocate, writers block and jams are symptoms not bugs to fix
46.
▲
by
marmaduke
2y ago
This is awesome! I was just today building a new wsl distro, and I didn’t want it tied to just a version of Linux but rather a project. I had to download a rootfs of Ubuntu and import it, and then spent time installing everything. Ofc I
47.
▲
by
marmaduke
2y ago
> stating the obvious What is obvious is wildly different among people. For instance, it was obvious to me that the article was sharing some ideas freely, and those ideas are ones which are not obvious to everyone in the workplace. “It’
48.
▲
by
marmaduke
2y ago
What kind of code did you try on the CPU for, say, ternary gemm? I imagine ternary values maps nicely to vectorized mask instructions, and much of tiling etc from usual gemm
49.
▲
by
marmaduke
2y ago
> CSCGs build on top of cloned HMMs by augmenting the model with the actions of an agent. pretty interesting take, and reminds me of the models that Karl Friston and many colleagues make under the umbrella term "active inference&quo
50.
▲
by
marmaduke
2y ago
The Human Brain Project's final 3 year period could have actually delivered a platform for actually modeling real human brain data, and GDPR totally blew that possibility out of the water: no one had the budget to take on the legal ri
51.
▲
by
marmaduke
2y ago
Can this run NFS or other network filesystems?
52.
▲
by
marmaduke
2y ago
C'mon, call it dark forest already
53.
▲
by
marmaduke
2y ago
I don't remember the title now, and it's hard to find them on YouTube these days.. but it's at the start of a section specifically on the diamond sutra. Maybe this one ? I will listen later to check https://youtu.b
54.
▲
by
marmaduke
2y ago
I like your summary. I also like Alan Watts' comment, that the mind is like a diamond, totally transparent but also the hardest, durable aspect of our existence.
55.
▲
by
marmaduke
2y ago
Very nice write up. Those are the kind of matrix sizes that MKL is fairly good at, might be worth a comparison as well? Also, if you were designing for smaller cases, say MNK=16 or 32, how would you approach it differently? I'm imple
56.
▲
by
marmaduke
2y ago
Yes, amen There's not having to exaggerate and moreover there's value in rehashing known but tricky stuff
57.
▲
by
marmaduke
2y ago
Ive dredged though Julia, Numba, Jax, Futhark, looking a way to have good CPU performance in absence of GPU, and I'm not really happy with any of them. Especially given how many want you to lug LLVM along with. A recent simulation cod
58.
▲
by
marmaduke
2y ago
The lack of review can be obviated by googling, tho it's exacerbated by lack of publishing in general for this domain, likely cause it's not work done at a classic academic department, and no motive to publish for citation /
59.
▲
by
marmaduke
2y ago
Yes! I have at least a few colleagues who do best coding on the train or airplane. I used to regularly switch coffee shops when switching tasks or projects. Async is great, but it is hard to do partially. Sync is infectious.
60.
▲
by
marmaduke
2y ago
Can't the cafe just charge by the hour, like a co-working space ?
More ›