박종민

@jijupax

세상의 활로에서!!

Yongin-si, Republic of Korea
Vrijeme pridruživanja: listopad 2009.

Tweetovi

Blokirali ste korisnika/cu @jijupax

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

  1. proslijedio/la je Tweet
    29. sij

    Good hyperparameters lead to ML success! 🙌 Learn how you can find the best parameters with Keras Tuner— a fully-featured, scalable, and easy-to-use hyperparameter tuning library for Keras and beyond. Read the blog →

    hyperparameters with keras tuner
    Poništi
  2. proslijedio/la je Tweet
    28. sij

    New paper: Towards a Human-like Open-Domain Chatbot. Key takeaways: 1. "Perplexity is all a chatbot needs" ;) 2. We're getting closer to a high-quality chatbot that can chat about anything Paper: Blog:

    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    26. sij

    Quaternions and Euler angles are discontinuous and difficult for neural networks to learn. They show 3D rotations have continuous representations in 5D and 6D, which are more suitable for learning. i.e. regress two vectors and apply Graham-Schmidt (GS).

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    24. sij

    To ensure new AI technologies will have a positive impact, we need to study societal implications from the beginning. Our new work in JAMA w/ of Harvard Law considers ethical and legal aspects of ambient intelligence in hospitals:

    Poništi
  5. proslijedio/la je Tweet
    23. sij

    Here's a lecture by Vivienne Sze () on efficient computing for deep learning, robotics, and AI as part of the MIT Deep Learning lecture series. Making real-world AI systems efficient is one of the most exciting and impactful problems in AI today.

    Poništi
  6. proslijedio/la je Tweet
    21. sij

    🚨New lecture series🚨 We've teamed up with to bring you the Deep Learning Lecture Series: 12 lectures covering a range of topics in Deep Learning - all led by DeepMind researchers, all free, and all open to everyone. Info & tickets:

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

    Want six free chapters and YouTube screencasts of the new Lite for Microcontrollers TinyML book? Check out !

    Poništi
  8. proslijedio/la je Tweet
    16. sij

    🗣Take an inside peek into the TensorFlow team’s own internal training sessions! 👀 In this episode of Inside TensorFlow, you’ll learn more about the TensorFlow Model Optimization Toolkit, particularly in quantization and pruning. Watch now →

    Poništi
  9. 10. sij

    Python 2의 EOL가 이유인거고 Python 3.5도 얼마 남지 않았으니 TF의 새로운 기능을 쓰고 싶다면 Python 3.6 이상을 권장

    Poništi
  10. proslijedio/la je Tweet
    9. sij

    2.1 is out 👍 Notable features: - The experimental TextVectorization layer allows you to include your text processing logic inside your model (for cleaner deployment & serialization) - The standard TF pip package now includes GPU support by default

    Poništi
  11. proslijedio/la je Tweet

    Our field isn't quite "artificial intelligence" -- it's "cognitive automation": the encoding and operationalization of human-generated abstractions / behaviors / skills. The "intelligence" label is a category error

    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet

    TensorFlow has a suite of tools for optimizing your models for faster inference: This includes post-training weight quantization, and gradual weight pruning during training for your Keras models.

    Poništi
  13. proslijedio/la je Tweet
    2. sij

    AugMix in TF2.0 1. Fully modular code 2. Custom train/eval loops with tf.function 3. **Custom EarlyStopping for the custom train loop.** 4. Checkpoint manager 5. Parallelized data generators

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

    Happy New Decade! In the last decade AI went from niche to mainstream. I wonder what our community will accomplish this decade?

    Poništi
  15. proslijedio/la je Tweet
    26. pro 2019.

    Starting Jan 6, we're doing a series of lectures at MIT on deep learning and AI. Skip the first one, but afterwards there are some great talks (inc. , ). All are welcome. Seating limited (1st come, 1st served). Video will be posted here:

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    24. pro 2019.

    Some folks still seem confused about what deep learning is. Here is a definition: DL is constructing networks of parameterized functional modules & training them from examples using gradient-based optimization....

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

    Excited to share our work on view synthesis! We generate new views of a scene from a single image. The model reasons about 3D structure without 3D supervision, trained end-to-end on image pairs Led by , with and Rick Szeliski

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet

    If you're a deep learning research looking for an introduction to what TF 2.0 has to offer, check out this notebook guide I made in October.

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet
    8. pro 2019.

    "maybe if you weren't on your cell phone all the time you'd be able to buy a house"

    Poništi
  20. proslijedio/la je Tweet

    Introducing : a new way to share your ML experiment results online! Read the blog →

    TensorBoard.dev
    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:

    ·