Erdan Genc

It's like when all the traps get re-opened in ghostbusters.

@midlandsound youtu.be/KnN2U0ZnBn4

This may revolutionize data science: we introduce TabPFN, a new tabular data classification method that takes 1 second & yields SOTA performance (better than hyperparameter-optimized gradient boosting in 1h). Current limits: up to 1k data points, 100 features, 10 classes. 🧵1/6

@ApacheUIMA I love the new setDataPathElements(File/String...) and getDataPathElements() methods! Thank you guys!

📣 Apache UIMA 3.3.0 has just been released! 🤩 Check out the release notes 👀 What's your favourite feature or fix? 🤔 github.com/apache/uima-ui… 💾 uima.apache.org/downloads.cgi #Apache #OpenSource #release

If you want to quickly prototype rule-based text annotation, we have a great new environment ready for you: IRuta - try it out now: mybinder.org/v2/gh/averbis/…

Ruta + Jupyter = 🧡 github.com/averbis/IRuta

Fine-tune any of the 11,000 🤗 models in Amazon SageMaker, And now, grab your custom code snippet right from the model page! 🔥🔥🔥 Learn more at hf.co/amazon

In 1998 you could fit the whole internet on one floppy disk.

Got tired of waiting to select which OS to boot, so I made a physical switch

We wanted to use Instagram ads to highlight how ad tech invades your privacy. Instead, Facebook shut our account down: signal.org/blog/the-insta…

@alpen_Q the matrix

How to get into woodworking: Step 1 - Learn to code Step 2 - Burn out Step 3 - Get into woodworking

"yeah, we automated that with a jupyter notebook"

bei inzidenzen von 100 aronal, ab 200 elmex

ccc.de/de/updates/202… ☝️

We're back this week with two new FSDL lectures. 1. Deploying machine learning models. 2. Monitoring models in production. We cover strategies, tools, and concrete recommendations for how to get your models into production and keep them there. Enjoy! fullstackdeeplearning.com/spring2021/lec…

What I’m really looking for when hiring an engineer twitter.com/matvelloso/sta…

Looking forward to present our #NLProc work on "Clinical Outcome Prediction from Admission Notes" at next week's #EACL2021 We deploy a new admission->discharge task setup & propose outcome pre-training to integrate knowledge from unlabeled data. Preprint bit.ly/3284fQM