andrewgross

7 ms·

1. 1.▲
   by andrewgross 2y ago

   Super easy to get started, but lacking for larger datasets where you want to understand a bit more about predictions. You generally lose things like prediction probability (though this can be recovered if you chop the head off and just assi
2. 2.▲
   by andrewgross 2y ago

   Got it. I’ll review the paper again for that portion. However, it still sounds like the end result is not VRAM savings but efficiently and speed improvements.
3. 3.▲
   by andrewgross 2y ago

   Ahh got it, thanks for the pointer. I am surprised there is enough correlation there to allow an entire GPU to be specialized. I'll have to dig in to the paper again.
4. 4.▲
   by andrewgross 2y ago

   Is there a concept of an expert that persists across layers? I thought each layer was essentially independent in terms of the "experts". I suppose you could look at what part of each layer was most likely to trigger together and s
5. 5.▲
   by andrewgross 2y ago

   > The beauty of the MOE model approach is that you can decompose the big model into a collection of smaller models that each know different, non-overlapping (at least fully) pieces of knowledge. I was under the impression that this was n
6. 6.▲
   by andrewgross 3y ago

   4 x 32GB for now. I need to investigate manual OC as EXPO doesn't work with all 4 slots populated. Another option is to try the 192GB 4x48GB Corsair kits at 5200.
7. 7.▲
   by andrewgross 3y ago

   7950x3d w/ 128GB at stock timings (~3200MT/s?). Showed a high base but no increase with threads, need to investigate what is happening. 1 54.7 2 50.6 3 49.6 4 48.2 5 47.9 6 47.4 7 47.1 8 46.6 9 46.5 10 46.2
8. 8.▲
   by andrewgross 12y ago

   Are you referring to the drive to use more of a languages' built in operators to build queries, instead of my approach? I think the Mozilla library uses an approach like that. http://elasticutils.readthedocs.org/en&#x2
9. 9.▲
   by andrewgross 12y ago

   It is a modified version of what we use at Yipit. I stripped out mostly things related to working with Django/Flask apps. It comes in very handy when making sure we maintain style and push fewer broken commits. Feel free to re-use.
10. 10.▲
    by andrewgross 12y ago

    Good point, I will have to swap out some testing libs but should be possible.
11. 11.▲
    by andrewgross 12y ago

    Thanks for sharing. Nice to see an alternative approach for defining these sorts of things.
12. 12.▲
    by andrewgross 12y ago

    Fixed. Completely missed that.
13. 13.▲
    by andrewgross 12y ago

    Thanks, checking this out.
14. 14.▲
    by andrewgross 12y ago

    Agreed. I had a rough time when I was first learning how to manually build queries as there are few examples of complex queries.
15. 15.▲
    by andrewgross 12y ago

    Thanks for sharing this, didn't know about it. Always nice to see how other people tackle the same issue.
16. 16.▲
    by andrewgross 12y ago

    We used to use Haystack, but found it a bit too opinionated for us once we wanted to do some custom stuff. It is a bit more faithful to the Django queryset API, something we had to abandon to let us use more of the complex Elasticsearch qu
17. 17.▲
    Show HN: PyEQS – query Elasticsearch like a Django Queryset(github.com)

    55 points by andrewgross 12y ago|24 comments
18. 18.▲
    OOP in Python (live streaming class for beginners) (enginehere.com)

    14 points by andrewgross 13y ago|2 comments
19. 19.▲
    by andrewgross 13y ago

    A friend of mine made NSArray extensions with a Ruby syntax a while back. Certainly useful if you find these helpful. https://github.com/nsantorello/nsarray-extensions
20. 20.▲
    by andrewgross 13y ago

    Awesome, thanks for the explanation
21. 21.▲
    by andrewgross 13y ago

    Thanks, was reviewing it with a coworker and crept up on the same conclusion. Seems a bit hard to test and definitely alarming at first glance.
22. 22.▲
    Ask HN: Can someone explain why this line exists?

    25 points by andrewgross 13y ago|16 comments
23. 23.▲
    by andrewgross 14y ago

    Not a bad idea since it would make easier to pull changes outside of the automation without needing to specify the GIT_SSH variable.
24. 24.▲
    by andrewgross 14y ago

    I know where you are coming from with compiled languages. A big reason for creating this flow was for working with repositories that are not our primary code repository. We have several internal tools and forks that don't have a complex d
25. 25.▲
    by andrewgross 14y ago

    Switch memcache to UDP for testing and see what it looks like, curious if TCP connection startups are causing issues.