Alex Amadori

@tacchinotacchi

Software Engineering student at 42, Paris. Indepentent AI/ML researcher. 🇮🇹🇬🇧🇯🇵🇫🇷

Vrijeme pridruživanja: prosinac 2012.

Tweetovi

Blokirali ste korisnika/cu @tacchinotacchi

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

  1. Prikvačeni tweet
    15. stu 2019.

    I'm excited to share this blog article on the project I've been working on for the past few months: Training a tiny Sentiment Analisys model using Knowledge Distillation to remarkably improve accuracy Article: Code:

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

    🇫🇷 42CTF est une plateforme de challenges propulsée par des étudiants de 🇺🇸 42ctf is a Capture The Flag platform powered by students 🏴‍☠️

    Poništi
  3. 18. pro 2019.

    Visualizing quaternions, by and These interactive videos are revolutionary! I wonder if something similar can be done to help explain Machine Learning

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

    Video of my 2nd talk. Starts at 30:00 here: Materials Matter: How biologically inspired alternatives to conventional neural networks improve meta-learning and continual learning. Slides:

    Poništi
  5. proslijedio/la je Tweet
    11. pro 2019.
    Poništi
  6. proslijedio/la je Tweet

    Self-supervised pre-trained models that can be used as base models for building 3D medical imaging models via transfer learning - - better than training 3D models from scratch and better than 2D approaches

    Poništi
  7. proslijedio/la je Tweet
    2. pro 2019.

    I've been working with for a long time to create something that combines the best of Notebooks with the best of traditional software development approaches. It's called nbdev. We're releasing it today as open source. 1/

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    18. stu 2019.

    It took a while, but I finally got the approval to release the implementation for the 3D Ken Burns project!

    Poništi
  9. proslijedio/la je Tweet
    18. stu 2019.

    They claim this system is robust, flexible. But can they really claim to be robust without a giraffe perturbation test? ಠ_ಠ "Reusable neural skill embeddings for vision-guided whole body movement and object manipulation." abs: pdf:

    Poništi
  10. proslijedio/la je Tweet

    There is no choice for those who cherish freedom but to support the protests in Hong Kong, as a bill pending in the United States Congress does

    Poništi
  11. proslijedio/la je Tweet
    15. stu 2019.

    Some of the smartest people I know are leaving AI research for engineering/neuroscience. Their reasons? 1. We need to understand how humans learn to teach machines to learn. 2. Research should be hypothesis -> experiments, but AI research rn is experiments -> justifying results.

    Poništi
  12. 16. stu 2019.

    Coded in and with 's transformers library!

    Prikaži ovu nit
    Poništi
  13. proslijedio/la je Tweet
    15. stu 2019.

    "From now on, you're not allowed to look at an equation (or especially type it in LaTeX) without also typing it in Python and actually calculating some values. Ideally, you should also plot some values" - p2 lec 10 "this is the only way to get

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    14. stu 2019.

    In early days of my career I made a postdoc's code go ~50 times faster. He promised co-authorship on the paper since his method was now viable. Months later, paper came out without my name. Why? Supervisor saw I only changed ~20 lines out of ~1000 and said was not significant.

    Prikaži ovu nit
    Poništi
  15. proslijedio/la je Tweet

    I've just released a fairly lengthy paper on defining & measuring intelligence, as well as a new AI evaluation dataset, the "Abstraction and Reasoning Corpus". I've been working on this for the past 2 years, on & off. Paper: ARC:

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    27. ruj 2019.

    Very interesting poll! Not many uni students I speak to who are trying to get that vital first job spend time on gh/kaggle. And few HR departments mention it in their job ads.

    Poništi
  17. proslijedio/la je Tweet
    17. ruj 2019.

    We've observed AIs discovering complex tool use while competing in a simple game of hide-and-seek. They develop a series of six distinct strategies and counterstrategies, ultimately using tools in the environment to break our simulated physics:

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet

    When haters go after your looks and differences, it means they have nowhere left to go. And then you know you’re winning! I have Aspergers and that means I’m sometimes a bit different from the norm. And - given the right circumstances- being different is a superpower.

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet

    Training a high-res generative model with a genetic algorithm

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    25. kol 2019.

    Who knew there was a sort of hanky code for hexadecimal. Thanks Datamation magazine from 1968

    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:

    ·