Lei Li

@_TobiasLee

Senior student at Xidian University, Homepage:

People's Republic of China
Vrijeme pridruživanja: kolovoz 2015.
0 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @_TobiasLee

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

  1. proslijedio/la je Tweet
    11. sij

    “Meet AdaMod: a new deep learning optimizer with memory” by Less Wright

    211 proslijeđenih tweetova 821 korisnik označava da mu se sviđa
    Poništi
  2. proslijedio/la je Tweet
    6. sij

    10 ML & NLP Research Highlights of 2019 New blog post on ten ML and NLP research directions that I found exciting and impactful in 2019.

    287 proslijeđenih tweetova 920 korisnika označava da im se sviđa
    Poništi
  3. 23. pro 2019.

    together，make Arsenal great again

    Mikel Arteta has confirmed that Freddie Ljungberg will stay at the club and be part of his backroom staff.
    1 korisnik označava da mu se sviđa
    Poništi
  4. proslijedio/la je Tweet
    23. pro 2019.

    New NLP News: 2020 NLP wish lists, HuggingFace + fastai, NeurIPS 2019, GPT-2 things, Machine Learning Interviews via

    1 reply 130 proslijeđenih tweetova 508 korisnika označava da im se sviđa
    Poništi
  5. 23. pro 2019.

    Still a lot to explore about GNN models

    paper on "Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological View" In this work, we present a systematic and quantitative study on the over-smoothing issue of GNNs.
    1 korisnik označava da mu se sviđa
    Poništi
  6. proslijedio/la je Tweet
    8. pro 2019.

    Our recent work to be presented at NeurIPS 2019: Understanding and improving layer normalization-- We find that the normalization effects on derivatives are much more important than forward normalization. Paper:

    1 reply 5 proslijeđenih tweetova 3 korisnika označavaju da im se sviđa
    Poništi
  7. proslijedio/la je Tweet
    5. pro 2019.

    A simple module consistently outperforms self-attention and Transformer model on main NMT datasets with SoTA performance. arxiv: code:

    3 proslijeđena tweeta 3 korisnika označavaju da im se sviđa
    Poništi
  8. proslijedio/la je Tweet
    26. stu 2019.

    Introducing the SHA-RNN :) - Read alternative history as a research genre - Learn of the terrifying tokenization attack that leaves language models perplexed - Get near SotA results on enwik8 in hours on a lone GPU No Sesame Street or Transformers allowed.

    The SHA-RNN is composed of an RNN, pointer based attention, and a “Boom” feed-forward with a sprinkling of layer normalization. The persistent state is the RNN’s hidden state h as well as the memory M concatenated from previous memories. Bake at 200◦F for 16 to 20 hours in a desktop sized oven.
    The attention mechanism within the SHA-RNN is highly computationally efficient. The only matrix multiplication acts on the query. The A block represents scaled dot product attention, a vector-vector operation. The operators {qs, ks, vs} are vectorvector multiplications and thus have minimal overhead. We use a sigmoid to produce {qs, ks}. For vs see Section 6.4.
    Bits Per Character (BPC) onenwik8. The single attention SHA-LSTM has an attention head on the second last layer and hadbatch size 16 due to lower memory use. Directly comparing the head count for LSTM models and Transformer models obviously doesn’tmake sense but neither does comparing zero-headed LSTMs against bajillion headed models and then declaring an entire species dead.
    550 proslijeđenih tweetova 1.933 korisnika označavaju da im se sviđa
    Poništi
  9. proslijedio/la je Tweet
    24. stu 2019.

    A simple module consistently outperforms self-attention and Transformer model on main NMT datasets with SoTA performance.

    19 proslijeđenih tweetova 53 korisnika označavaju da im se sviđa
    Poništi
  10. proslijedio/la je Tweet
    8. stu 2019.

    slides from my keynote talk at

    91 proslijeđeni tweet 366 korisnika označava da im se sviđa
    Poništi
  11. proslijedio/la je Tweet
    5. stu 2019.

    This slide from 's (excellent) @emnlp2019 talk speaks to me on a spiritual level. From: Show Your Work: Improved Reporting of Experimental Results. Jesse Dodge, Suchin Gururangan, Dallas Card, Roy Schwartz and Noah A. Smith Paper link:

    7 proslijeđenih tweetova 52 korisnika označavaju da im se sviđa
    Poništi
  12. proslijedio/la je Tweet
    26. ožu 2019.

    New blog post (...and new blog) on how to build a deep learning framework from scratch in 100 LOC . Let me know what you think!

    52 proslijeđena tweeta 161 korisnik označava da mu se sviđa
    Poništi
  13. 27. ožu 2019.

    Congrats!

    I am extremely honored to be the recipient of the 2018 ACM A.M. Turing Award, and absolutely delighted to be sharing it with my friends and colleagues Geoffrey Hinton and Yoshua Bengio. NYT:...
    Poništi
  14. 22. pro 2018.

    great summarization~

    10 Exciting Ideas of 2018 in NLP: A collection of 10 ideas that I found exciting and impactful this year—and that we'll likely see more of in the future.
    1 korisnik označava da mu se sviđa
    Poništi
  15. proslijedio/la je Tweet
    27. stu 2018.

    + are the fastest platform for ML. On the popular (and now quite silly) ResNet-50 ImageNet benchmark, one TPUv3 Pod finishes in 2.2 minutes at full accuracy (76.3% top-1)! Approx 2x faster than extremely large GPU clusters. Details:

    110 proslijeđenih tweetova 309 korisnika označava da im se sviđa
    Poništi
  16. proslijedio/la je Tweet
    26. stu 2018.

    "Pattern Recognition and Machine Learning" by is now available as a free download. Download your copy today for an introduction to the fields of pattern recognition & machine learning:

    21 reply 1.616 proslijeđenih tweetova 2.879 korisnika označava da im se sviđa
    Poništi
  17. proslijedio/la je Tweet
    11. lis 2018.

    A new era of NLP has just begun a few days ago: large pretraining models (Transformer 24 layers, 1024 dim, 16 heads) + massive compute is all you need. BERT from : SOTA results on everything . Results on SQuAD are just mind-blowing. Fun time ahead!

    13 replies 474 proslijeđena tweeta 1.109 korisnika označava da im se sviđa
    Poništi
  18. 4. srp 2018.

    funny

    This new family of GAN loss functions looks promising! I'm especially excited about Fig 4-6, where we see that the new loss results in much faster learning during the first several iterations of training. I implemented the RSGAN loss on a toy problem and it worked well.
    1 korisnik označava da mu se sviđa
    Poništi
  19. 31. lis 2017.

    Long time no see! I'm online again just for twitter dataset QAQ

    1 korisnik označava da mu se sviđa
    Poništi
  20. 1. kol 2015.

    怕什么真理无穷，进一寸有一寸的欢喜。 Hello,Twitter!

    2 proslijeđena tweeta 9 korisnika označava da im se sviđa
    Poništi

@_TobiasLee 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:

    ·