Loreto Parisi

@loretoparisi

Computer Engineer MSc. Technical Director of Machine Learning Father of an amazing girl and a cute boy.

Bologna, Italy
github.com/loretoparisi
Vrijeme pridruživanja: kolovoz 2007.
1.586 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @loretoparisi

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

  1. proslijedio/la je Tweet
    prije 19 sati

    New models from: - (Dutch BERT), - at Facebook AI (MMBT, multi-modal model) - , et al. (FlauBERT, French-trained XLM-like) - , et al. at (UmBERTo, Italian CamemBERT-like)

    , , i još njih 3
    1 reply 8 proslijeđenih tweetova 34 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  2. 2. velj

    “Compared to an existing state-of-the-art generative model, OpenAI , has 1.7x greater model capacity and was trained on 8.5x more data.” The way to generative text models seems to be more data and more capacity. But are we sure it’s true?

    1 reply 4 proslijeđena tweeta 4 korisnika označavaju da im se sviđa
    Poništi
  3. 1. velj

    Hey guys look, is here! welcome! 🐍

    Hello, world! This is the official account for NumPy, the fundamental package for scientific computing with Python. Follow us for news, info and content related to NumPy!
    2 korisnika označavaju da im se sviđa
    Poništi
  4. proslijedio/la je Tweet
    1. velj
    Odgovor korisnicima

    4 proslijeđena tweeta 8 korisnika označava da im se sviđa
    Poništi
  5. 31. sij

    Kudos to for his work on text generation with LaserTagger. 🚀

    Finally it's out! A blog post giving an overview and some background on our work on text generation with LaserTagger.
    2 korisnika označavaju da im se sviđa
    Poništi
  6. proslijedio/la je Tweet
    31. sij

    I already find their RoBERTo and BERT model useful for English. They now release for other languages. 👏 - UmBERTo, a Roberta-based Language Model trained on large Italian Corpora. - FlauBERT, A new French model was added. - Dutch BERT model.

    Transformers 2.4.0 is out 🤗 - Training transformers from scratch is now supported - New models, including *FlauBERT*, Dutch BERT, *UmBERTo* - Revamped documentation - First multi-modal model, MMBT from , text & images Bye bye Python 2 🙃
    Prikaži ovu nit
    1 reply 7 proslijeđenih tweetova 28 korisnika označava da im se sviđa
    Poništi
  7. 30. sij

    Fast, parallel (browser) applications with SIMD

    WebAssembly brings SIMD support to the Web. What is it? and explain and demonstrate how to use it to run native code on the Web even faster! 🔥 (This feature is currently experimental and available under a flag.)
    1 proslijeđeni tweet 1 korisnik označava da mu se sviđa
    Poništi
  8. 29. sij

    Let’s support core-js, a modular standard library for and that includes polyfills for . Made by

    4 proslijeđena tweeta 3 korisnika označavaju da im se sviđa
    Poništi
  9. 29. sij

    List of English contractions from in

    Poništi
  10. proslijedio/la je Tweet
    24. sij

    is now using adjacent sentences as context! I mentioned on the TAUS call that I had noticed DeepL seems to be doing it. Just now showed me an example that confirms it. Beautiful to see this is now reality.

    i
    5 proslijeđenih tweetova 12 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    28. sij

    I'm happy to announce our latest work on self-supervised learning for . PASE+ is based on a multi-task approach useful for recognition. It will be presented at . paper: code: @Mila

    7 replies 82 proslijeđena tweeta 302 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  12. 27. sij

    The heroines Grace Hopper (1906-1992) was a trailblazing computer programmer who became the first person to call a programming problem named after she found a dead moth 🐜 in a malfunctioning computer

    2 korisnika označavaju da im se sviđa
    Poništi
  13. 25. sij

    Demystifying computing

    The world’s top experts explain why Quantum computing is so important
    Poništi
  14. proslijedio/la je Tweet
    22. sij

    So, a blog post. Not connected to Marian at all, surprisingly, but about a MT quality rabbit hole I went down the last couple of days. "Is MT really lexically less diverse than human translation?" If that's useful in any way I might keep on blogging.

    9 replies 21 proslijeđeni tweet 60 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  15. 25. sij

    accuracy by theory and in practice it’s not the same thing. This work could fill the gap!

    This has surprisingly little effect on prediction accuracy, but does improve the match between networks considered by theory and used in practice. With Roman Novak, , and . Implementation in .
    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    24. sij

    We're releasing mBART, a new seq2seq multilingual pretraining system for machine translation across 25 languages. It gives significant improvements for document-level translation and low-resource languages. Read our paper to learn more:

    15 replies 390 proslijeđenih tweetova 1.052 korisnika označavaju da im se sviđa
    Poništi
  17. 24. sij

    Wow! audio models in essentia audio library. That’s cool congrats

    Two papers accepted to !! "An empirical study of Conv-TasNet", my first paper with . "Tensorflow audio models in Essentia", one of my last papers with the ! This is a nice connection for an ICASSP that is going to be held in Barcelona! :)
    1 reply 1 korisnik označava da mu se sviđa
    Poništi
  18. 24. sij

    Easy audio waveform generation in

    4 proslijeđena tweeta 2 korisnika označavaju da im se sviđa
    Poništi
  19. proslijedio/la je Tweet
    3. pro 2019.

    We've just updated our paper on Stacked Capsule Autoencoders -- an unsupervised model that achieves SOTA classification performance with , and Fig: TSNE plot of capsule activities on MNIST.

    33 proslijeđena tweeta 160 korisnika označava da im se sviđa
    Poništi
  20. proslijedio/la je Tweet
    22. sij

    Looking for a reference or example of using the SentencePieceBPETokenizer in the new tokenizers library. Any suggestions ?

    3 proslijeđena tweeta 13 korisnika označava da im se sviđa
    Poništi

@loretoparisi još nije objavio nijedan Tweet.

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:

    ·