dimitry12

8 ms·

1. 61.▲
   by dimitry12 8y ago

   Thanks! I will consider trying paid version.
2. 62.▲
   by dimitry12 8y ago

   Cool! What are they? Have you considered adding the feature of listing the conversations which contributed to the score?
3. 63.▲
   by dimitry12 8y ago

   @lettergram In lettergram.com, when it says "{} opinion ({}) in {} separate discussions" - are these discussions from Reddit and HN?
4. 64.▲
   Why YC “office hours” rock(blog.pletnikov.com)

   1 points by dimitry12 8y ago|0 comments
5. 65.▲
   by dimitry12 8y ago

   I am impressed the number of answers though!
6. 66.▲
   by dimitry12 8y ago

   @lainon now I am curious why you would ask such a question :D What are you going to do with the answers?
7. 67.▲
   by dimitry12 8y ago

   Some fields on the "COMPANY DETAILS"-form are obviously too short. For example "How far along are you?" is 60 characters.
8. 68.▲
   by dimitry12 8y ago

   I don't but I was researching the space and https://www.anodot.com/ has the most feature-rich product - though they only discover anomalies in numeric time-series.
9. 69.▲
   by dimitry12 8y ago

   Are they doing supervised training for anomalies?
10. 70.▲
    Unit-tests: solution for combinatorial explosion of black-box tests(blog.pletnikov.com)

    2 points by dimitry12 8y ago|0 comments
11. 71.▲
    by dimitry12 8y ago

    Both require model tuning before you can get a speed-up from mixed-precision training.
12. 72.▲
    by dimitry12 8y ago

    It was in stock on and off and I was able to order it directly from Nvidia US. After 59 days of playing with it, I sent it back (initiated return on 30th day, after I already figured out it doesn't live up to the hype, then had another
13. 73.▲
    by dimitry12 8y ago

    An important hidden cost here is coding a model which can take advantage of mixed-precision training. It is not trivial: you have to empirically discover scaling factors for loss functions, at the very least. It's great that there is n
14. 74.▲
    by dimitry12 9y ago

    Both Machine Learning (ML) and Probabilistic Programming (PP) work towards building a mathematical object (model), which takes observation as input and produces prediction as output. Both ML and PP are about finding the parameters of the mo
15. 75.▲
    by dimitry12 10y ago

    Can anybody recognize the headset they are using? Or is it Disney's completely in-house device?
16. 76.▲
    by dimitry12 10y ago

    Instapaper is down in US as well. I agree with your surprise.
17. 77.▲
    by dimitry12 12y ago

    Company: http://www.unicity.com/usa/company/ - Unicity International located in Orem, UT. Position: Sr Software Engineer. REMOTE and VISA are cool. I'm Software Architect here and I'm looking for a perso
18. 78.▲
    by dimitry12 15y ago

    I'm in. Got the 3rd ed. while waiting for the shipment of the 4th (possibly long wait). Let me know if you're in a similar situation.
19. 79.▲
    by dimitry12 17y ago

    The scenario is: keep my 1 TB of data and store backup completely off-site, without having local copy of it when doing increments (so the backup tool can not compare the original data and backup to find the changes, but need to consult cat
20. 80.▲
    Ask HN: Offsite backup, BUT without local copy of the data (with catalog)?

    3 points by dimitry12 17y ago|3 comments