brrrrrm

14 ms·

151.
by brrrrrm 3y ago
OMP_NUM_THREADS=n is pretty standard in HPC workloads
152.
Is Elon specially promoting Mr Beast's recent video?(twitter.com)
2 points by brrrrrm 3y ago|1 comments
153.
by brrrrrm 3y ago
Whoa thanks!
154.
by brrrrrm 3y ago
PeachPy is awesome - it's really a great way to play with assembly-level performance programming and not have to deal with assembler toolchains
155.
by brrrrrm 3y ago
I haven't found a library that makes me feel comfortable with using WebRTC for random little projects. Setting up TURN/STUN is annoying and I've actually never successfully setup a server/client model (it seems like it&
156.
by brrrrrm 3y ago
I work professionally in C++, but I wouldn’t tout it either. I find Python to be the most pleasant personally.
157.
by brrrrrm 3y ago
Weird to think typescript is similar to a language that requires a class definition for “hello world” lol. You can’t even fit it on a single 80 character line…
158.
by brrrrrm 3y ago
yea, I guess it's the lack of human-friendly default assumptions mixed with prescriptive design patterns. definitely an anachronistic design intent :(
159.
by brrrrrm 3y ago
it feels like the characters-typed to functionality ratio is quite low. at the same time, the code isn't much cleaner and there's a lot of mental overhead in deciphering the abstractions (in this case Collector).
160.
by brrrrrm 3y ago
Isn’t it run multiple times and the slowest is dropped? Why would you expect the file to be evicted from ram so quickly?
161.
by brrrrrm 3y ago
looking at the sample code I’m quite glad I’ve never really touched Java. What a clunky language
162.
Drawer Drawer (draw by clicking drawers)(draw-her-a-drawer.glitch.me)
2 points by brrrrrm 3y ago|0 comments
163.
by brrrrrm 3y ago
matmul is old and useful -- there's a lot of hardware on a chip that makes it run much faster (prefetch, vectorization, instruction parallelism) and some of these languages have optimizations to expose those things automatically.
164.
by brrrrrm 3y ago
doesn't look that easy... zend_parse_parameters? pre-baked configure + make scripts? check out how a modern language deals with this stuff https://bun.sh/docs/api/ffi#usage
165.
Show HN: Brr.js – Trying to make WebGPU a bit easier to use(github.com)
2 points by brrrrrm 3y ago|0 comments
166.
by brrrrrm 3y ago
whoops, I forgot to include the URL. dang would you mind deleting this?
167.
[forgot to include URL, won't let me add it]
1 points by brrrrrm 3y ago|1 comments
168.
by brrrrrm 3y ago
great line: "cynicism masquerades as wisdom"
169.
by brrrrrm 3y ago
as documented it looks different (async vs sync + the necessity of `text:` annotations in the Node.js version?) updated my comment to paste in what's written in the docs
170.
by brrrrrm 3y ago
why on earth did they design the Node.js and Web APIs to be slightly different and incompatible? (edit: this might just be a bug/oversight on the landing page?) Node.js: const model = genAI.getGenerativeModel({ model: "gemin
171.
by brrrrrm 3y ago
they should make it accessible at https://imagen.google like how meta did with https://imagine.meta.com
172.
by brrrrrm 3y ago
mixture of experts gates on the feed forward network only. the shared weights are the KQV projections for the attention mechanism of each layer.
173.
by brrrrrm 3y ago
I once met a Google PM whose job was to manage “Easter eggs” in the Google home assistant. I wonder how many engineers effectively “hard coded” features into this demo. (“What the quack” seems like one)
174.
by brrrrrm 3y ago
factorization methods are somewhat uncommonly used in deep learning (the likely target of this framework) and have compute properties (such as approximate outputs, non-deterministic number of iterations) that make them unlike the BLAS++ sta
175.
by brrrrrm 3y ago
> 1. They can get equivalent position and pay at the new Microsoft startup during that time, so their jobs are not at risk. citation?
176.
by brrrrrm 3y ago
this was announced last week. The major changes are memory capacity and bandwidth. That means smaller batch sizes (i.e. inference) will be faster (1.4-1.7x) and larger batch sizes (i.e. training) will be possible. Current large batch siz
177.
by brrrrrm 3y ago
can I try this out somewhere?
178.
by brrrrrm 3y ago
Sustainability is a luxury belief. The market will change when such values disseminate more broadly.
179.
by brrrrrm 3y ago
ML has extremely predictable and heavily optimized routines. Languages that can target hardware ISA all tend to have comparable perf and there’s no reason to think Rust would offer much.
180.
by brrrrrm 3y ago
Genuine question: have macro trends ever, in the history of humanity, been consistently and correctly analyzed by fine-grained time series data? I was always a bit sketched out by the macro economics classes I took in college.

More ›