Elijah 'bewpie' Rippeth

@terrible_coder

Extremely unprincipled engineer and linguist at . I cannot solve problems with computers. All opinions my own. he/him

Washington, DC
Vrijeme pridruživanja: kolovoz 2014.

Tweetovi

Blokirali ste korisnika/cu @terrible_coder

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

  1. Prikvačeni tweet
    Poništi
  2. 5. velj
    Poništi
  3. proslijedio/la je Tweet
    5. velj

    *sips some good beer* *looks at the time* hmm lets see its about 8:55 PM *ponders my impending doom*

    Poništi
  4. 4. velj

    putting your computer programs online is posting cringe

    Poništi
  5. proslijedio/la je Tweet
    3. velj
    Poništi
  6. 3. velj

    Turns out he's 100% A White Guy who doesn't pay taxes

    Poništi
  7. 2. velj

    just made my first contribution to ignite as I work toward this (lofty? arduous?) goal. 🙀

    Prikaži ovu nit
    Poništi
  8. 2. velj

    hoping to refactor fairseq to use ignite for training, metrics, and checkpointing... dogfooding the pytorch ecosystem

    Prikaži ovu nit
    Poništi
  9. 1. velj

    Thinking about barbe à papa

    Poništi
  10. 1. velj

    if I were a PhD student, I would simply produce an error-free dissertation

    Poništi
  11. 29. sij
    Poništi
  12. proslijedio/la je Tweet
    28. sij
    Odgovor korisniku/ci

    Fortnite player said

    Poništi
  13. 26. sij

    Rest in peace, black mamba

    Poništi
  14. 26. sij

    By the time the exception is thrown, the file handle has been closed because the file owning object has been GC'd. 🙀

    Prikaži ovu nit
    Poništi
  15. 26. sij

    Got some wild stuff happening on Windows where with TemporaryDirectory("foo") as dir_: foo(dir_) is throwing an exception because `foo` hasn't released references to dir_ by the time it returns.

    Prikaži ovu nit
    Poništi
  16. 25. sij

    A pre-nup, but for ssh keys.

    Poništi
  17. 22. sij

    Heuristic approach to bootstrap for training data: return true

    Prikaži ovu nit
    Poništi
  18. 22. sij

    Free idea: reply-guy classifier

    Prikaži ovu nit
    Poništi
  19. 22. sij

    I think might be a good option where the train task is some background task. does this seem reasonable?

    Prikaži ovu nit
    Poništi
  20. 22. sij

    Does anyone know of some asynchronous serve/train web servers for machine learning models? I'd like to be able to train some light sklearn models while serving a current model - almost like blue-green deployments.

    Prikaži ovu nit
    Poništi
  21. 21. sij
    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:

    ·