kettleballroll

13 ms·

91.
by kettleballroll 5y ago
That sounds fairly childish. Aside from the very obvious fact that you'll miss out on a lot of cool stuff, if you don't agree with the value judgements of your peers, the problem most likely resides with your choice of peers.
92.
by kettleballroll 5y ago
off-topic, but: Go and Rust seem to aim for the same space (system programming), so when is one preferable to the other, in your opinion?
93.
by kettleballroll 5y ago
Bayesian optimization and bayesian data analysis are two different things. Your suggestion seems off topic to me, and it feels like you didn't even check the linked content before posting.
94.
by kettleballroll 5y ago
I can't really any good ways for them to raise money, outside of becoming a subscription service or selling user data. What ways did you imagine? Because the more I think about it, the harder this problem appears to be...
95.
by kettleballroll 5y ago
> has already resulted in large-scale leaks of sensitive information I must have missed that, could you share more?
96.
by kettleballroll 5y ago
That makes sense, thanks for elaborating! I could imagine a different solution where instead you just don't report the overall star-average, to make it a harder metric to game, but still get to have a stronger feedback signal for your
97.
by kettleballroll 5y ago
Do you have any evidence for that claim, or is that pure speculation on your part?
98.
by kettleballroll 5y ago
Pretty well? Sure. Win a competition? No. If you're not running standard models (where you only have learning rate and regularization to grid over) but you're trying more specialized stuff/developing new models, 4 hyperparame
99.
by kettleballroll 5y ago
I disagree. I've participated and even won a few of those "ML in sciences" competitions in the past, and all other things being equal, the one with the most compute wins. 12$/hour sounds doable, but as soon as you need t
100.
by kettleballroll 5y ago
At these sizes, tpu would definitely be the way to go, and would likely be a lot cheaper (and potentially faster) than GPUs.
101.
by kettleballroll 5y ago
In a previous life, almost a decade ago, I fought very similar fights with OpenMP and MKL using R. It's painful and you need to pay heed to all these small details pointed out in the docs as in OPs case. However, it's worth noting
102.
by kettleballroll 5y ago
Awesome improvements, congrats!
103.
by kettleballroll 5y ago
> in this case it takes 4-8 threads or so for KySync to match Zsync's performance. Assuming zsynch is single threaded, in an apples/apples comparison (1 thread) zsynch wins by over 10x. I can't help but wonder how much fas
104.
by kettleballroll 5y ago
As opposed to which other company?
105.
by kettleballroll 5y ago
This is are already happening. eg in Germany, young Friday-for-the-future activists won their court case arguing that currently lenient climate goals are unfair to future generations ( https://www.theguardian.com/world&#x2F
106.
by kettleballroll 5y ago
It's always nice to see big labs working more towards building an understanding of things instead of just chasing SOTA. On the other hand, I'm not sure there is a lot of actionable findings in here. I guess that's the trade o
107.
by kettleballroll 5y ago
You're being cryptic. The people your criticizing (this might include me, i have no way of knowing) will not be able to learn anything from your reply. It would be more constructive I'd you shared your thoughts so we'd have a
108.
by kettleballroll 5y ago
As far as I know, hogwild doesn't really scale to very large systems and doesn't get used in practice, but maybe that's just because it's most useful on CPUs, which are inefficient for other reasons
109.
by kettleballroll 5y ago
I think it would be interesting to have progressive decoding (and, for those who want it, autoplay) of videos, too.
110.
by kettleballroll 5y ago
> whatever will eventually become the standard, will be lossless, as it became with audio At least ony bubble, the standard for audio is still lossless: I'm fairly sure 99.9% of all audio i consume is using something lossy:streamed
111.
by kettleballroll 5y ago
What does his religion have to do with his ability to be far-right?
112.
by kettleballroll 5y ago
Wit seems to me like this failure waa decades in the making anyways. What would you have done differently if you were in Biden's shoes?
113.
by kettleballroll 5y ago
You make it sound like this is Netflix's fault, whereas tlit seems to me that 10 years ago it was mich easier to acquire licenses for streaming. now some of the big studios have their own streaming services, so licensing isich harder&#
114.
by kettleballroll 5y ago
Master/Slave nomenclature is considered "problematic language" (because it harks back to slavery), and thus often replaced with more PC language these days. My organization had a whole "fix this nomenclature throughout t
115.
by kettleballroll 5y ago
OTOH, everyone dreams of building a successful SaaS, because they want to milk that model themselves.
116.
by kettleballroll 5y ago
It sounds like the people that are being made redundant were mainly tasked with labeling data (which also matches them being temps instead of full time employees, and them training replacements in Costa Rica), so I don't think they&#x2
117.
by kettleballroll 5y ago
Am an academic too. I don't find this to be common knowledge, but a common misconception. Would like to see a citation
118.
by kettleballroll 5y ago
> in the past few years that this is very difficult nowadays. I strongly disagree. It has always been as difficult as today (if not more), it's just that these days were aware of it.
119.
by kettleballroll 5y ago
> Were all dancing around the elephant in the room that clearly everyone panicked due to china's lies at the start of 2020 Given what I've witnessed since the start of the pandemic in my own country, it seems more plausible to