Congrats to Joshua Fan, Junwen Bai, Carla Gomes (@CornellCIS) & Zhiyun Li (Cornell Dyson) for winning the best paper award (ML innovation category) at the Climate Change AI #NeurIPS2021 workshop! Very proud of our students! 🍾🎉

Congrats to our #NeurIPS2021 workshop best paper winners in the category "ML innovation"🏆 Joshua Fan, Junwen Bai, Zhiyun Li, Ariel Ortiz-Bobea & Carla Gomes (all Cornell University) for work on spatio-temporal GNNs for crop yield prediction. Well deserved! 👏🌱🤖 climatechange.ai/papers/neurips…

In this week's Deep Learning Paper Review, our researchers examine Joint Unsupervised and Supervised Training For Multilingual #ASR. Read on for our key takeaways 👇 assemblyai.com/blog/review-ju… #DeepLearning #SpeechRecognition

Announcing FlashAttention, a fast and memory-efficient attention algorithm with no approximation! 📣 w/ Dan Fu By reducing GPU memory reads/writes, FlashAttention runs 2-4x faster & requires 5-20x less memory than PyTorch standard attention, & scales to seq. length 64K. 1/

CoDA brings parameter efficiency and inference efficiency simultaneously. The idea derives from NLP and extends to Speech & Vision with impressive results!