ovi256

69 ms·

691.
by ovi256 9y ago
>While mosh provides the same core functionality as ET, it does not support native scrolling nor tmux control mode (tmux -CC).
692.
by ovi256 9y ago
Regarding your prediction of a music streaming service, Prime Music and Music Unlimited already exist! https://www.amazon.com/gp/dmusic/promotions/PrimeMusic https://www.amazon.com/gp/dmu
693.
by ovi256 9y ago
Hmm I would refrain from saying DL is useful only when approaching 1 megapixel images. State of the art performance on MNIST is held by a 6 layer convnet (4 layers convolutional, 2 layers FC). MNIST is just 28 x 28 grayscale images, so 768
694.
by ovi256 9y ago
It's a very narrow ruling, conditioned on the fact that the defendant, Hancom, publicly acknowledged they were using Ghostscript under GPL. It does not say much, if anything, but IANAL, about a defendant who would not have acknowledged
695.
by ovi256 9y ago
It's not, for the project you have to setup a Google Cloud account, so you have the same restrictions like everyone else: 60 min of free service per month, then $0.006 per 15s.
696.
by ovi256 10y ago
To help others understand why, the "Deep Learning" book, which summarizes current state-of-the-art DL, advises having at least 5000 samples (images) per class for OK performance (equal to non-DL approaches) and 50k to 100k for sta
697.
by ovi256 10y ago
I'm sure it's buried somewhere in the Terms of Use. Contemporary terms of use are a classic dark pattern. If one would like to setup an useful ML project, take as input Terms of Use, spit out the highlights, the buried ledes, such
698.
by ovi256 10y ago
What you read on the public site is the USP for the users that install the unroll.me app. They are not the economic buyers, the ones who pay for the service. Those are the people that buy the user's aggregate data. They get in-person p
699.
by ovi256 10y ago
This is a case of developper ignorance, not missing tools: v8 has an integrated debugger since 2009, and you can interact with it using the Chrome developer tools, probably the most used debugger frontend of all. There are also debuggers fo
700.
by ovi256 10y ago
Except the razor is $400 here. It's not a throwaway device like a razor or a printer.
701.
by ovi256 10y ago
A Chromebook would be the only non-crap new machine for that budget. You can install Linux on Chromebooks.
702.
by ovi256 10y ago
This is about hitting the high notes, as Joel Spolsky put it in his [post]. A lot of people are good software engineers and can wrote good software. To get Jeff Deans that write revolutionary software, we don't yet have a better method
703.
by ovi256 10y ago
That would be very interesting because it would remove all risk for the founders of being considered "US persons", which has unexpected and unpleasant consequences such as not being accepted as a client by many non-US banks or fin
704.
by ovi256 10y ago
Lawyers like faxes because it leaves a legally acknowledged written record on both sides, and they're instantaneous. So they send a lot of them. For some reason the fax is legally sounder than email.
705.
by ovi256 10y ago
Of course this type of warming is not dangerous on a human lifetime timescale. It would also require us to become a level I Kardashev Scale civilization (harnessing all the energy reaching our planet from our star, or the equivalent). But i
706.
by ovi256 10y ago
>No more global warming. Au contraire, my friend. Unlimited cheap energy would bring on, in time, the next type of anthropogenic warming: the temperature of the planet being raised, not by a greenhouse effect, but by the raw power output
707.
by ovi256 10y ago
An Uber driver who gets a single passenger complaint of not having been sober is reviewed. If there are police records such as a DUI ticket to confirm the complaint, he's excluded. This is better than occasional testing.
708.
by ovi256 10y ago
Mandating the use of an obsolete technology (taxi meter, when a smartphone + app + backend has the same capability) seems like sensible regulation to you ? That regulation was tailor-made to shut down Uber operations, as a favour to the tax
709.
by ovi256 10y ago
"If you give me six lines written by the hand of the most honest of men, I will find something in them which will sell millions of pageviews." There, updated for contemporary KPIs.
710.
by ovi256 10y ago
I see what you mean by overbetting on one idea, but please stop selling this point that they're throwing their lives away. That's unnecessary hyperbole. Even if nothing works at all, which is rarely the case, because the BATNA is
711.
by ovi256 10y ago
The 15 Mbps is TCP bandwidth, and the 100 Mbps is physical layer bitrate. They are not comparable. There are overheads and latency and packet loss can reduce how much physical bitrate TCP can use. So the 4k stream may or may not do great in
712.
by ovi256 10y ago
This will enable writing a plagiarism detector which will not be fooled by the simple strategy of replacing words with their synonyms. Given that synonyms have very similar embeddings, you can compute a distance between two phrases by compu
713.
by ovi256 10y ago
The military already uses anti-radiation missiles with home-on-jam capabilities extensively. Here's the US model, which will carry a 66kg HE warhead to 150 km: https://en.wikipedia.org/wiki/AGM-88_HARM
714.
by ovi256 10y ago
If you're in Paris, go meet people from recruiting agencies like Mobiskill or Urban Linker. It's literally their job to know the market and find opportunities for you. In one or two emails you would be getting great advice about m
715.
by ovi256 10y ago
OVH.com is selling 2GB VMs for 2.99 Euros per month. No idea about over-provisioning or any comparison with Linode though.
716.
by ovi256 10y ago
Another great example is the last post on the keras blog [1] "Using pre-trained word embeddings in a Keras model". You take advantage of a large pre-trained network for a text classification task. The OpenFace face recognition lib
717.
by ovi256 10y ago
Sure, but the most interesting interpretations are usually based on aprioris learned in this universe.
718.
by ovi256 10y ago
Khan Academy is an amazing true MOOC resource. The calculus goes up to early undergrad. The differential calculus is the most important for ML, in my oppinion. Probability and statistics will help ML as well. I would recommend Khan Academy
719.
by ovi256 10y ago
Do I even need to point out this is a 3D rendering ? One of the creators, Vitaly Bulgarov, is a well-known concept designer and animator, as extensively documented on his websites. He sells 3D designs and training materials. He has worked o
720.
by ovi256 10y ago
The phenomenon you describe, demand elasticity, exists just as well in the medical market. There's a lot of rationing going on, explicit or implicit. It's just that the elasticity coefficient may be lower compared to the smartphon

More ›