abhishek thakur

@abhi1thakur

Chief Data Scientist @ . First 3x GM on Kaggle: . Co-Organizer of Berlin ML Meetup. ML Comic.

Vrijeme pridruživanja: prosinac 2009.

Tweetovi

Blokirali ste korisnika/cu @abhi1thakur

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @abhi1thakur

  1. Prikvačeni tweet
    30. pro 2019.

    It seems I made it to Top-5 highest rank in Kaggle in all categories in 2019! Yay! :)

    Prikaži ovu nit
    Poništi
  2. prije 5 sati

    I reached 5000 subscribers on YouTube and I have a special something to offer y'all!! Check it out here:

    Poništi
  3. 3. velj

    yep. most of the people dont care about what degree you have as long as you understand stuff and can implement it.

    Poništi
  4. 2. velj

    Scrolling through some Kaggle Kernels sometimes makes me sleepy because of all the messiness. Everything becomes even more messy when one has to work in teams. In my new video, I show how to create much cleaner inference kernels. Check it out here:

    Poništi
  5. 1. velj

    Figuring out correct cuda and cudnn versions for new Tensorflow versions. 🤣🤣🤣

    Poništi
  6. 1. velj

    One step closer to become the first 4xGM on Kaggle. This step is gonna take a while because upvotes for datasets are really difficult to earn.

    Poništi
  7. 1. velj

    Have you ever wondered how to make a clean submission Kernel on kaggle's kernel competitons? My next episode will be dedicated to this and we will learn how to make a super clean inference kernel/notebook on Kaggle!!! Subscribe Now!!!

    Poništi
  8. 31. sij

    Microsoft CEO: Indian Alphabet CEO: Indian IBM CEO: Indian India and Indians rule!!!!

    Poništi
  9. 30. sij

    Also, suggestions for episodes, tips n tricks and special episodes are most welcome. So, please let me know in this tweet thread what you want to see and I'll make it happen!

    Prikaži ovu nit
    Poništi
  10. 30. sij

    It's been a month and my YouTube channel now has 4700+ subscribers and 25K+ total views! whts coming next: - entity embeddings - live competiton coding - ensembling - docker - and much more!! (not in this order) If u havnt subscribed yet, SUBSCRIBE NOW!!

    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    29. sij

    This Monday, in class, explained how the retina performs a 100× compression to go from 100M photoreceptors to 1M axons in the optic nerve. As a consequence, the central object in this video appears to become brighter / darker although its brightness is constant. – mjesto: NYU Center for Data Science

    Poništi
  12. 27. sij

    BERT on Steroids!!! This special episode on my YouTube channel explains how you can finetune a BERT model on a classification/regression datasets using PyTorch and Google Cloud TPUs!!! Check it out here:

    Poništi
  13. 25. sij

    BERT on steroids: Finetuning BERT for a dataset using pytorch and TPUs!

    Poništi
  14. 23. sij

    forgot to add: in production. 🤣

    Prikaži ovu nit
    Poništi
  15. 23. sij

    Upgrading python from 3.5 to 3.7 is not a piece of cake. Understood the hard way today.

    Prikaži ovu nit
    Poništi
  16. 22. sij

    Good that I already know about these scams and how they work.

    Prikaži ovu nit
    Poništi
  17. 22. sij

    How is it possible that within a month of installing windows 10, I got a call from windows support scammer ? I installed and used windows after 7 years and in those 7 years I didnt get any calls like that. Are insiders providing this information?

    Prikaži ovu nit
    Poništi
  18. 21. sij

    How to fine-tune BERT using PyTorch on Google Cloud's TPUs? Coming on my YouTube channel this week! Stay tuned!

    Poništi
  19. 21. sij

    Are there any plans to support python 3.5 with tokenizers?

    Poništi
  20. 20. sij

    Today, i deactivated my Facebook. Tired of people.

    Poništi
  21. 19. sij

    The first part of Handling Categorical Features in my video series is now live!!! Check it out here:

    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:

    ·