Shubham Chandel

@sksq96

Grad student at NYU working in NLP and machine learning. Previously IBM research, Amazon.

New York, NY
Vrijeme pridruživanja: rujan 2011.
Rođen/a 1996.

Tweetovi

Blokirali ste korisnika/cu @sksq96

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @sksq96

  1. Prikvačeni tweet
    17. lip 2018.
    Poništi
  2. proslijedio/la je Tweet
    30. sij

    We're standardizing OpenAI's deep learning framework on PyTorch to increase our research productivity at scale on GPUs (and have just released a PyTorch version of Spinning Up in Deep RL):

    Poništi
  3. proslijedio/la je Tweet
    23. sij
    Odgovor korisnicima

    I propose Ravana, a mythological 10-headed figure as the closest thing to a mascot for transformers:

    Poništi
  4. proslijedio/la je Tweet
    21. sij

    Universals of word order reflect optimization of grammars for efficient communication: a thread. Joint work with and

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    21. sij

    Many thanks to a tip from , I've added a brief section to my self-supervised learning post on noise contrastive estimation and consistency loss. And thanks to I've also added a paragraph about the history of self-supervised learning

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    19. sij

    Ok, I got a pretty good port of these Maxwell equation figures into Python. Heavy use of JAX and . I'll tidy up the code post it here later.

    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet

    Earlier this year, I had a productivity breakthrough... I used to spend $25k-$100k building an app over 3-6 months. It was frustrating, expensive, and slow. Then I started using NoCode tools like Webflow, Bubble, Zapier and AirTable...

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    4. sij 2018.
    Odgovor korisnicima

    i’ve thought up a similar one before: happiness is the derivative of quality of life w.r.t. time. this explains why happiness seems invariant to absolute quality of life, wealth, etc. from monarch to peasant, you’re happy if you’re improving whatever it is you have

    Poništi
  9. proslijedio/la je Tweet
    10. sij

    A nice article on the interplay between physics and machine learning, which focuses specifically on geometric deep learning and gauge equivariant NN. Featuring , Miranda Cheng, and myself.

    Poništi
  10. proslijedio/la je Tweet
    7. sij

    I completed my 1st data science project ~30 years ago. Since then I've been continuously developing a questionnaire I use for all new data projects, to ensure the right info is available from the start. I'm sharing it publicly today for the first time.

    Poništi
  11. proslijedio/la je Tweet

    NIPS 1990 had papers about "hybrid symbolic-connectionist" models. This is now as far in the past as 2050 is in the future.

    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    1. sij

    Wishing everyone a happy, healthy and productive year! Don’t worry about research trends and the current publication frenzy—dig into hidden corners, explore intellectual traditions others are ignoring, and learn new things, no matter your age and stage. And, be kind to yourself!

    Poništi
  13. proslijedio/la je Tweet
    1. sij

    Python 2 is now deprecated.

    Poništi
  14. proslijedio/la je Tweet
    31. pro 2019.

    “Technical writing gave me back the good feeling of working with text. At the same time, it forced me to streamline a bunch of thoughts I had over months in text. This transformed thoughts into arguments that can be reused and shared.” —

    Poništi
  15. proslijedio/la je Tweet
    25. pro 2019.
    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    26. pro 2019.

    I'm spending too much time writing about ML drama. I need to spend more time writing about sillier things instead. Anyways, here's a post about NeurIPS.

    Poništi
  17. proslijedio/la je Tweet
    20. pro 2019.

    Always remember, outcome = effort * conditions. This is an important lesson to keep in mind when working in other high stress situations like startups, for eg. Being just outcome focused will not produce the right compassion for the conditions/context.

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet

    I make models I don’t understand of a thing I don’t understand, using algorithms I don’t understand, on compute platforms that at least other people understand.

    Poništi
  19. proslijedio/la je Tweet
    17. pro 2019.

    Not everyday evaluation is the focus of attention as in this survey (). If you are unfamiliar with the issues, start here, and start now! But don't expect to end there (See comments).

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    17. pro 2019.

    Here's my conversation with Michael Stevens (), creator of the amazing Vsauce YouTube channel. His videos explore questions and ideas in science that educate, entertain, and inspire millions of people. Also, he's just fun to talk to:

    Poništi
  21. proslijedio/la je Tweet
    16. pro 2019.

    My 2020 NLP wishlist -start paying attention to human lang acquisition, to learn more about us & gain insights for NLP models -for similar reasons, delve deeper into lang evolution & emergence -recover from the recent community-wide amnesia & remember there was NLP even before DL

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·