New code walkthrough on https://t.co/m6mT8SrKDD: image search with natural language queries using a dual-encoder approach. The idea is to learn a joint embedding space for images and associated captions (inspired by OpenAI CLIP).https://t.co/fWHsv18QuL
— François Chollet (@fchollet) January 26, 2021
GPU accelerated feature engineering on @Kaggle? Yes Please! Check out our latest post showing how you can use NVTabular and @RAPIDSai to speed up workflows by 11x and training of @fastdotai on Kaggle by 2x in this great blog post by Benedikt Schifferer: https://t.co/4FK4xA2Rij
— Even Oldridge (@Even_Oldridge) January 26, 2021
#ipygany 0.5.0 is out!
— Martin Renou (@martinRenou) January 26, 2021
The Python library for scientific visualization in a @ProjectJupyter Notebook, powered by your GPU!
This release adds support for JupyterLab 3 and new colormaps and colorbars.
Repo: https://t.co/L1qkg790Lq
Live docs: https://t.co/OKTgd5tIPj pic.twitter.com/K0mGYCCJRu
When reading a paper I can skim through a lot by just asking 5 simple questions:
— Eric Jang 🇺🇸🇹🇼 (@ericjang11) January 26, 2021
1 What are the inputs ?
2 What are the outputs ?
3 What loss supervises the output predictions and what assumptions about the world does this make?
Such wildly differing standards for air quality depending on the country. Air that’s considered Very Poor in the #EU and Unhealthy in the #USA is considered Good in #China. #dataviz
— Randy Olson (@randal_olson) January 25, 2021
Source: https://t.co/9ucRZuqiUs pic.twitter.com/WVrFhjO7Ce
I'm with @fchollet on this. There're some best-practices on creating and organizing data that experienced applied ML people use, but we still need to flesh out and widely disseminate these ideas. This will be key to getting more ML systems deployed. https://t.co/wJQEwNU6SX
— Andrew Ng (@AndrewYNg) January 25, 2021
Rich people think they are middle class. pic.twitter.com/I0xxhlWEST
— John B. Holbein (@JohnHolbein1) January 25, 2021
📑 For anyone who is getting started with user experience or product management in the machine learning space: the PAIR guidebook is an excellent, open-source resource:https://t.co/Y9SEWhgmfW
— 👩💻 Paige Bailey @ 127.0.0.1 🏠 (@DynamicWebPaige) January 25, 2021
👥 Main ideas: interacting with users, asking the right questions, & defining metrics pic.twitter.com/qJk5UWjoxH
Statisticians: We assume the data are independent.
— Christoph Molnar (@ChristophMolnar) January 25, 2021
The data: https://t.co/mBnT535WDa
ML researchers work with fixed benchmark datasets, and spend all of their time searching over the knobs they do control: architecture & optimization. In applied ML, you're likely to spend most of your time on data collection and annotation -- where your investment will pay off.
— François Chollet (@fchollet) January 24, 2021
Chart shows which generation controls the US Senate. Forget all the talk about millennials, it’s boomer time! Source: https://t.co/NOS4evOS2v pic.twitter.com/gP9g0Kw3K4
— Simon Kuestenmacher (@simongerman600) January 23, 2021
“You can’t just flip a switch and be a software company.” https://t.co/OC8du4MscC
— hardmaru (@hardmaru) January 23, 2021
