vonnik

62 ms·

631.
by vonnik 10y ago
Curious: When did you play with DL4J and what kind of cluster are you running TF on? You're right to say those are two separate things, so let's compare apples to apples.
632.
by vonnik 10y ago
We're missing the "In" on "In" Q Tel in the title. And I have to add, IQT serves a ton of govt agencies, not just the CIA.
633.
by vonnik 10y ago
Fwiw, we're testing a Dockerized the distro of DL4J. Runs on DCOS. https://imgur.com/a/CDTAc https://imgur.com/a/6jlxi We'll release in coming weeks.
634.
by vonnik 10y ago
Hi Trevor - Nice to see you here. :) For data parallelism, we have a simple wrapper that covers as many GPUs as you want in one box. For more than one box, we use Spark, and the integration is explained here: http://deeplearning4
635.
by vonnik 10y ago
The environment for your large clusters is almost certainly different from everyone using TF outside of Google. I'm speaking about the problems they'll run into.
636.
by vonnik 10y ago
Done!
637.
by vonnik 10y ago
I love this piece. I also feel pessimistic about calm tech because a lot of tech companies have perverse incentives to addict and perpetually disturb their users in order to gain mindshare and sell ads.
638.
by vonnik 10y ago
Tensorflow is actually pretty slow and problematic on large clusters outside the Google Cloud. Probably because that's not what it was designed for. For Java/Scala people, Deeplearning4j has a pretty sophisticated Spark + GPUs set
639.
by vonnik 10y ago
The vast majority of Tensorflow users are students from the Udacity course. So the "rising dominance" of Tensorflow is actually disputable. 95% of those users have little experience, let alone the opportunity of deploying to produ
640.
by vonnik 10y ago
I find Russell's assumption that machines will learn good behavior from humans to be incredibly naive. History is a nightmare from which we cannot escape, as Joyce said. A robot exposed to news and general human behavior today would
641.
by vonnik 10y ago
Let's blame the math. It couldn't possibly be the fault of businesses targeting the poor.
642.
by vonnik 10y ago
There is literally no new information in this article about DL and FPGAs.
643.
by vonnik 10y ago
Katie Benner has written some good stories, but I'm suspicious of articles like this, because they insist almost irrationally on finding a trend that they can make a pronouncement about. The truth is, startups are all over the place. S
644.
by vonnik 10y ago
Reminds me of JavaCPP: https://github.com/bytedeco/javacpp
645.
by vonnik 10y ago
I don't think the article proves the point it claims to make (you don't need a founder), or disproves the message from YC: having a co-founder increases the likelihood that your startup will survive. First of all, most startups fa
646.
by vonnik 10y ago
Yeah, I probably plug DL4J too much. It's a hard habit to break, and I'm sure some people get tired of it. I try to only do it when I have something substantive to say about the post in question. OpenAI and DL4J are equivalent in
647.
by vonnik 10y ago
Anyone interested in this should look at JavaCPP, the bridge between Java and C++. https://github.com/bytedeco/javacpp Created and maintained by a Skymind engineer.
648.
Credentialism and Corruption: Vile College Administrators Edition(nakedcapitalism.com)
6 points by vonnik 10y ago|0 comments
649.
by vonnik 10y ago
You would need to join a community of people who can show you how to contribute to these projects. There are a lot of very active open-source communities, including OpenAI and Deeplearning4j, to name but two.[0][1] People in those communiti
650.
by vonnik 10y ago
Have you tried Yodlee? They serve as the API for many banks.
651.
by vonnik 10y ago
We've compiled a fairly extensive list of deep learning papers here: http://deeplearning4j.org/deeplearningpapers We also Tweet out new ones as they're published here: https://twitter.com/deeplear
652.
by vonnik 10y ago
I have a big problem with articles like this. Apple's PR is notorious for cracking the whip, which means that the "inside story", if they give it to you, comes with a warning to the journalist to behave and be nice. Levy&#x27
653.
by vonnik 10y ago
One worry here is that AI only classifies as well as the labeled datasets that it trains on. And my concern would be the way humans apply labels to other humans. Diagnosing mental illness is notoriously messy. DSM-IV descriptions of various
654.
by vonnik 10y ago
To paraphrase Bjarne Stroustrup: "There are two kinds of startups: the ones people criticize, and the ones nobody's ever heard of."
655.
by vonnik 10y ago
Glad you like the looks of it. It was built and is maintained by a Skymind engineer. All of us are on this Gitter channel if you have questions: https://gitter.im/deeplearning4j/deeplearning4j
656.
by vonnik 10y ago
Total tangent: Anyone doing scientific computing work on the JVM may be interested in: ND4J: N-dimensional arrays for the JVM http://nd4j.org/ Libnd4j: The C++ engine powering the above https://github.com/de
657.
by vonnik 10y ago
VCs' content marketing posts are the bleating edge of technology. That said, there are some good ones. They are usually written by long-time operators who flipped to investing. Or from VCs with a real sense for data analysis, combined
658.
by vonnik 10y ago
If anyone is interested in RNNs for Java and Scala, we've documented them here: http://deeplearning4j.org/usingrnns
659.
by vonnik 10y ago
The problem with news articles like this is that they attempt to appeal to a general readership through grand promises and fear-mongering. The reporter's basic problem is: How can I make my audience care? So you call up Nick Bostrom an
660.
by vonnik 10y ago
That's not true, actually. There's a ton of work being done in computational creativity. AI is actually pretty good at sensing similarities between instances of data, and recombining various elements to create something new. There

More ›