Rezultati pretraživanja
  1. 2. kol 2019.

    "Postering" participation in the worshop on translation of biomedical abstracts

  2. 1. kol 2019.

    Has MT reached human parity? Great discussion at

  3. 10. srp 2019.

    Wish to extract word alignment from NMT models without external aligners? Wondering whether NMT models learn interpretable word alignment beyond attention? Check out our new paper "Saliency-driven Word Alignment Interpretation for NMT" at co-located with !

    Prikaži ovu nit
  4. Facebook AI models achieved 1st place in 4 different language tasks of this year’s machine translation competition.

  5. 31. lis 2018.

    Language pairs for news task: English <-> Chinese, German, Gujarati and Russian. Hoping to confirm 1-2 more. Also aiming for unidirectional test sets, and document context in evaluation.

    Prikaži ovu nit
  6. 1. kol 2019.

    The official results of are out! Very proud of our () results! Our systems for English↔️Lithuanian achieved a tied first place!

  7. 27. srp 2019.

    On the train from Rome to Florence for ! Presenting work on Saliency-driven Interpretation of NMT at on Aug 1st. Also feel free to chat with me about GPU-friendly structured prediction, optimal BPE choice for low-resource NMT, or artwork/coffee in Florence!

  8. 22. sij 2019.

    Excited to announce the official "Machine Translation Robustness" task at ! Train your MT systems to be robust to lexical variation, slang, etc. using a real dataset from social media:

    Prikaži ovu nit
  9. Facebook AI has achieved superhuman performance on English to German translation tasks, according to organizers. Read more:

  10. 1. srp 2019.

    paper by Denis Emelin on "Widening the Representation Bottleneck in Neural Machine Translation with Lexical Shortcuts". shortcuts consistently improve Transformer - our analysis indicates that they free up representation capacity. with

  11. 3. sij 2019.

    🎤🎧🎼 🎸🎹🥁We are back in 2019 with an even more exciting program ! More to come in the coming weeks -

  12. 1. kol 2019.

    WMT 2019 - Possible low-resource languages datasets

  13. 2. kol 2019.

    Antonio Lopes so busy in describing the mystry behind the best APE system in the world!

  14. 2. kol 2019.

    Remarks from the Parallel Corpus Filtering task at

  15. 17. tra 2019.

    The best systems of each task , Cheers

  16. 2. kol 2019.

    delivering a thought provoking keynote at on semantic and stylistic divergences in parallel text.

  17. 2. kol 2019.

    presenting the findings of the APE shared task in at

  18. 13. tra 2019.

    We make to realize both genders that we need to work as a team. Balance is the thing that's needed.

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.