I want to retract this tweet publicly:
https://t.co/2iwQ8SoJT1
It is incorrect and causes unnecessary harm to the authors of "PoolFormer: MetaFormer is Actually What You Need for Vision" (https://t.co/iCewozc85R).
Using just AvgPool and MLP, they outperform most models.
1/6
It's week 12 of Machine Learning Zoomcamp!
We cover deploying deep learning models with AWS Lambda
🔸 Introduction to Serverless
🔸 TensorFlow Lite
🔸 Preparing Lambda code
🔸 Preparing a Docker image
🔸 Creating the lambda function
🔸 API Gateway
👉 https://t.co/o2sNjRaCB3
Since the GNNs for missing data are trending, I feel inclined to advertise our method published 2018 (https://t.co/uW5bgAZZn6) which learns, for each node and feature type, a vector for the available and one for the missing values. These are aggregated during message passing.
Pre-training large models is useful, but adds many hyperparameters. E.g. task weights, or augmentations in SimCLR. We give a scalable, gradient-based way to tune these:
https://t.co/Gv2luOFHxL
https://t.co/KTD4BM6ZTk
w/ @RaghuAniruddh@jonLorraine9@skornblith@MattBMcDermott
Congratulations to #CoRL2021 best paper finalist, "Robot Reinforcement Learning on the Constraint Manifold", Puze Liu, Davide Tateo, Haitham Bou Ammar, Jan Peters.
https://t.co/SdssOu3akD
#robotics#learning#award#research
Chengdu Supercomputing Center has been put into use for one year. With 100000 trillion operations per second, the host performance ranks among the top ten in the world. .
Mainland China’s capital market will triple in size to US$100 trillion in next 10 years - SCMP
https://t.co/lSURsdxbn7
How competitive is China’s economy compared to the US? Check out our feature on China’s global competitiveness: https://t.co/ltBSihhA6i
We have been developing and open-sourcing parts of a touch-sensing ecosystem. This week, we announced that GelSight will be manufacturing DIGIT, a high-resolution, compact tactile sensor designed for robotic in-hand manipulation. Learn more: https://t.co/nQP8ND3X6z