3abiton

30 ms·

1. 541.▲
   by 3abiton 2y ago

   Isn't this what a MoE LLM does already?
2. 542.▲
   by 3abiton 2y ago

   > This AI bubble could be burst by a better cheaper product more dedicated to the task True, but currently, no one can match the offering of NVidia, they are the sole pioneer in this market. AMD is yet to show any viable option (both on
3. 543.▲
   by 3abiton 2y ago

   Any benchmarks on performance for on-demand embeddings?
4. 544.▲
   by 3abiton 2y ago

   More like not totally baseless rumors. https://www.tomshardware.com/tech-industry/qualcomm-faces-be...
5. 545.▲
   by 3abiton 2y ago

   But not same performance. One of the reason QC not releasing anything but controlled benchmarks is most likely the subpar performance of Windows on ARM. This will be the biggest hurdle for QC Elite chips, competing with M-seried which is de
6. 546.▲
   by 3abiton 2y ago

   What's the relevance compared to the latest AI advancement including transformers for example?
7. 547.▲
   by 3abiton 2y ago

   I did not know that python fact. Thanks for sharing!
8. 548.▲
   by 3abiton 2y ago

   But why Clojure, I am curious about the choice of the language?
9. 549.▲
   by 3abiton 2y ago

   But why would be better than Mathematica for example for calculus?
10. 550.▲
    by 3abiton 2y ago

    3blue1brown made also great tools for math visualizations, his work is so impactful in the educational YouTube domain.
11. 551.▲
    by 3abiton 2y ago

    Good boi AI. GoodboAI?
12. 552.▲
    by 3abiton 2y ago

    This is and will be a rallying moment soon for the community, both open hardware and software finally working together! This will be huge by the end of the decade.
13. 553.▲
    by 3abiton 2y ago

    Which one would be best fitting a replacement for Azure DevOps?
14. 554.▲
    by 3abiton 2y ago

    Does it relate to applications in time series? How does the hierarchy play a role in Transformers?
15. 555.▲
    by 3abiton 2y ago

    So buying them separately and mixing them in your mouth is legal? It's like combining 2 pieces of open-source software and selling them under one product.
16. 556.▲
    by 3abiton 2y ago

    Any real world performance comparison between QLoRa and LoRa?
17. 557.▲
    by 3abiton 2y ago

    Only Nvidia? These predictions are maybe misleading? And might present bad information to unaware users?
18. 558.▲
    by 3abiton 2y ago

    > That's a 57x speedup by doing nothing. Okay, I'll have what you're having.
19. 559.▲
    by 3abiton 2y ago

    This problem specifically (3D reconstruction with representation fitting) is really an overfitting nightmare, they just adapted to it not really overcame it. Nonetheless interesting work.
20. 560.▲
    by 3abiton 2y ago

    I am not sure if "good rule of thumb" is a good standard of comparison. It's at best anecdotal.
21. 561.▲
    by 3abiton 2y ago

    How does it differ from OpenDashboard?
22. 562.▲
    by 3abiton 2y ago

    Can you patent open-source code?
23. 563.▲
    by 3abiton 2y ago

    You were not kidding, it's much more detailed than I anticipated down to the physics theory of they work.
24. 564.▲
    by 3abiton 2y ago

    What is the difference?
25. 565.▲
    by 3abiton 2y ago

    And data quality. Ensuring the sourcing and quality is very important to get a good model.
26. 566.▲
    by 3abiton 2y ago

    It describes the problem in a nicer way tbh. I haven't yet read the preprint but the video is neat.
27. 567.▲
    by 3abiton 2y ago

    Iam curious about the quantification of "super fast", do you have some pointers?
28. 568.▲
    by 3abiton 2y ago

    Kudos to nix again!
29. 569.▲
    by 3abiton 2y ago

    This also explains it's popularity and notoriarity.
30. 570.▲
    by 3abiton 2y ago

    Thanks for sharing your work!