Vitaly retweeted
Entropy is not the easiest thing to understand.
It is rumored to describe something about information and disorder, but it is not clear why.
What do logarithms and sums have to do with information?
💡 Let me explain! 💡
„Разрушим «Крепость». Вернём задержанным право на защиту” - Jetzt unterschreiben! chng.it/NzhzcTpr via @ChangeGER
A pretty unique find, a comprehensive and super useful guide to preparing for a coding interview.
freecodecamp.org/news/coding…
Finally someone is trying to attract attention to this problem 📰🇩🇪🌍
Bundesjustizministerin Christine Lambrecht: Stoppt ungewollte Werbepost! Schluss mit der Ressource... chng.it/57t2TdgY via @ChangeGER
Vitaly retweeted
Happy to announce telesto.ai, our project to make machine learning crowdsourcing production-ready! We are releasing our platform early and would like to continue the development with community feedback! (1/8)
Vitaly retweeted
Being resource-constrained forces you to ship solutions that do less, or do it in a simpler way. Often, these solutions turn out to be more effective than the over-engineered solutions shipped by resource-rich teams.
Resource constraints can be a design advantage.
Vitaly retweeted
Tip: excalidraw.com is fantastic for quickly mocking up UIs. Free, feature-filled and the live collaboration mode is excellent.
Great work by @Vjeux and team!
Thank you @ossia for writing this helpful article.
Learn to Code From Home: The Coronavirus Quarantine Developer Skill Handbook
freecodecamp.org/news/corona…
Vitaly retweeted
We've made a draft of our deep learning book available as @ProjectJupyter notebooks.
Important: This is for personal use, and not licensed for any redistribution or change of format or medium. Eg. don't make a PDF and share it, please!
github.com/fastai/fastbook
Vitaly retweeted
Glad to see Catalyst success in both @kaggle and @NeurIPSConf – 4th place on Recursion Cellular Image Classification!
When it comes to reproducibility, teamwork and fast R&D – Catalyst is a clue.
writeup: kaggle.com/c/recursion-cellu…
code: github.com/ngxbac/Kaggle-Rec…
Vitaly retweeted
Our paper on training with pseudo-labels for semantic segmentation, GCPR 2019.
Semi-Supervised Segmentation of Salt Bodies in Seismic Images:
SOTA (1st place) at TGS Salt Identification Challenge.
🌐 github.com/ybabakhin/kaggle_…
📝 arxiv.org/abs/1904.04445
#kaggle #TGS2019 #GCPR19
Vitaly retweeted
Look how folks at @getqonto are using DVC to version inputs, intermediate files and algorithm models💪
Reading that DVC is helping to 'drastically increase productivity' is making us so happy!🎯👏
medium.com/qonto-engineering…
International Summer School on Deep Learning is over. That was an amazing week thanks to the awesome speakers @apaszke @rasbt @alfcnz @alxndrkalinin and excellent organization. @muhakabartay thanks for the great networking ;)
#DLISS19 #gdansk
The third day of International Summer School on Deep Learning.
Ready to learn about image super resolution with #DeepLearning
Gaining more expertise for the challenging projects in @alexandrovteam
@instagram hi guys, how can my email already be used and not available for password recovery at the same time? Thanks!
Vitaly retweeted
#AI and #deeplearning people! So, recently I implemented a #CNN from scratch, essentially building a mini DL framework. Would anyone be interested if I were to expand how to build a general DL framework with modular components as a series of Jupyter notebook lessons?
Vitaly retweeted
Catalyst 19.06 has been released!
Lots of improvements:
- distributed training with fp16 support
- updated segmentation models contrib
- on-policy reinforcement learning
Read the release notes here ↓
github.com/catalyst-team/cat…