New work! Humans appear to learn similarly for different modalities and so should machines! data2vec uses the same self-supervised algorithm to train models for vision, speech, and nlp.
— Michael Auli (@MichaelAuli) January 20, 2022
Paper: https://t.co/bfcfP3dO1C
Blog: https://t.co/vi52LniOfv
Code: https://t.co/EVODCxOehP pic.twitter.com/98x0UxA44O
I just published "Stan Algorithms: Where to Start?" @mcmc_stan https://t.co/2Ui3kuBP1Q
— Daniel Lee (@djsyclik) January 19, 2022
Yeah, I can not only recommend one tool but two :). https://t.co/MjRedslvsB (lets you make AlexNet- and LeNet-style figures), and there is the https://t.co/BtNpVWzbDe repo, which looks amazing. Have fun!
— Sebastian Raschka (@rasbt) January 18, 2022
A nice thread that explains intuitive why self-attention mechanism is needed in order to properly represent the relationships of the inputs—something that MLP’s struggle with. Looking forward to the next thread! https://t.co/s5Zrl6cjbH
— hardmaru (@hardmaru) January 18, 2022
tplot - A library for creating text-based graphs in the terminal. https://t.co/lRcJYlSI3q #python pic.twitter.com/JJDAkJsQpN
— Python Weekly (@PythonWeekly) January 17, 2022
A Thousand Words Are Worth More Than a Picture:
— AK (@ak92501) January 17, 2022
Natural Language-Centric Outside-Knowledge Visual Question Answering
abs: https://t.co/pmciI9NUSg
TRiG framework outperforms all sota supervised methods by at least 11.1% absolute margin pic.twitter.com/CZkWbL4CPO
A good summary of various mental models useful for thinking about complex problems in many different areas, from science, systems, economics, to warfare.https://t.co/UHvjAWG7YE
— hardmaru (@hardmaru) January 17, 2022
TIL: in IPython, to insert a new line and turn a long Python oneliner expression into a more readable multi-line code snippet, you can press the following sequence:
— Olivier Grisel (@ogrisel) January 14, 2022
ctrl-q ctrl-j
Pressing "Enter" would instead have triggered the REPL execution, even with surrounding parens. pic.twitter.com/oDD186JR5x
GradMax: Growing Neural Networks using Gradient Information
— AK (@ak92501) January 14, 2022
abs: https://t.co/IyAuQffXcS
a new approach for growing neural networks that aims to maximize the gradient norm when growing pic.twitter.com/hDml0jMCbP
BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations
— AK (@ak92501) January 14, 2022
abs: https://t.co/bZ22mqHXEL
project page: https://t.co/hWQaDqkANr pic.twitter.com/SRH7XIIys1
We 💙 Grandmaster @Rob_Mulla's new #PogKaggle Community Competition series! Get started with these resources 👇
— Kaggle (@kaggle) January 14, 2022
📌 Join the Competition: https://t.co/SFc9ES5g7P
📌 Follow Rob on Twitch: https://t.co/puXSGAfvIK
📌 Check out his YouTube walk-through: https://t.co/HH17rhLqAX
Good discussion about ConvNets vs Transformers https://t.co/MmPInthvFH
— hardmaru (@hardmaru) January 13, 2022
