There are tons of things you can make with R Markdown at different points on the spectra of relevance and complexity. @seankross' postcards package aims to be high relevance and low complexity for students and folks new to R Markdown. https://t.co/ZHQLtYFjgj #cascadiaRconf
β Kara Woo (@kara_woo) June 5, 2021
When Vision Transformers Outperform ResNets without Pretraining or Strong Data Augmentations
β AK (@ak92501) June 4, 2021
pdf: https://t.co/GYknaVoNAM
abs: https://t.co/kaUxIdMVNQ
+5.3% and +11.0% top-1 accuracy on ImageNet for ViT-B/16 and MixerB/16, with the simple Inception-style preprocessing pic.twitter.com/EI1ZSUccUn
Cloud TPU Virtual Machines finally released
β hardmaru (@hardmaru) June 2, 2021
These VMs run on TPU host machines that are directly attached to TPU accelerators, so everything feels like it's run locally.
The new πΉπΆπ―ππ½π library supports TensorFlow, PyTorch, JAX, and soon, Julia. π₯https://t.co/Yhatj1xaKz pic.twitter.com/N6u3ix0MI0
You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection
β AK (@ak92501) June 2, 2021
pdf: https://t.co/LThmFQ2a6g
abs: https://t.co/XhhOip5bOw
github: https://t.co/crirLeiGGI
series of object detection models based on the naΓ―ve Vision Transformer pic.twitter.com/Ml38kzqdtt
π And merged to Transformers!
β Hugging Face (@huggingface) June 1, 2021
We are excited to welcome ByT5 as the first tokenizer-free model!
πAll available checkpoints can be accessed on the π€hub here: https://t.co/WEpfzu6uMN
π Demo (on master): https://t.co/2qgmt7YQO8 pic.twitter.com/m5pXe9qETU
StyTr^2: Unbiased Image Style Transfer with Transformers
β AK (@ak92501) June 1, 2021
pdf: https://t.co/H3OraPsolh
abs: https://t.co/PM8dZiCuct pic.twitter.com/8ld0m4SDyN
An Attention Free Transformer
β AK (@ak92501) June 1, 2021
pdf: https://t.co/iOURQMubTR
abs: https://t.co/6TSsVXmjww
an efficient variant of Transformers that eliminates the need for dot product self attention pic.twitter.com/ZfaIbdmvnL
Less is More: Pay Less Attention in Vision Transformers
β AK (@ak92501) June 1, 2021
pdf: https://t.co/ydo2bFvxsH
abs: https://t.co/baTSDrBpEd
hierarchical vision transformer pays less attention in
early stages to ease huge computational cost of self-attention modules over high-resolution
representations pic.twitter.com/6J5xdAO0mc
The quickest route along primary roadways to Washington D.C. from any point in the contiguous United States. #travel #dataviz
β Randy Olson (@randal_olson) May 31, 2021
Source: https://t.co/135uBbkwFu pic.twitter.com/T5yPHImNQl
The way I would improve spreadsheets is by allowing row numbering to start from any integer (including negative). Then row numbers could count what it makes sense to count. Books have done this forever via Roman numerals for front matter. Surely this isnβt so hard to do in 2021? pic.twitter.com/WMopl7uihp
β Christopher Manning (@chrmanning) May 29, 2021
Animated demographic pyramid of #Sweden from 1860-2020. #dataviz
β Randy Olson (@randal_olson) May 28, 2021
Source: https://t.co/ndxA55dzNI pic.twitter.com/I8an9Q05sg
What gripes do you have with LaTeX's default, and what you always add to papers? Here are mine: 1. Cleveref. Don't use "Section \ref{sec:intro}". Use \Cref{sec:intro}. This makes writing less error prone and it makes "Section" part of the hyperlink! https://t.co/SbJkDXJwjK
β Dustin Tran (@dustinvtran) May 28, 2021
