A Simple Convergence Proof of Adam and Adagrad https://
bit.ly/3zwWvvZ
#AI #MachineLearning #DeepLearning #LLMs #DataScience
MACHINE LEARNING
-
Simple Convergence Proof of Adam and Adagrad Optimizers
By
–
-
Four Data Trends Making An Impact On Marketing Now
By
–
Four Data Trends Making An Impact On Marketing Now
#AI #AIio #BigData #ML #NLU #Futureofwork @chrismessina @ChristopherIsak @davidwkenny @debashis_dutta @petitegeek @fabiomoioli @GaryMarcus @asokan_telecom http://
ow.ly/9AqQ30sH8ZO -
Google Research@ Event Showcases AI Progress and Academic-Industry Collaboration
By
–
Google Research's Research@ event in NYC convened 500+ leaders from academia, industry & startups to highlight the breadth & impact of Google's research. It featured progress across domains & fostered collaboration within the community. Were you there? Share your key insights ↓
-
Rachel Metz Interviews Nobel Laureate John Jumper on AlphaFold
By
–
I am SO EXCITED to be interviewing John Jumper, newly minted Nobel laureate and Google DeepMind director (and team leader of AlphaFold) on stage in London next Tuesday. Tune in to https://t.co/W38RELZmSq to check it out! (Also live on the Terminal!) https://t.co/7r4yGoCrTH
— Rachel Metz (@rachelmetz) 18 octobre 2024I am SO EXCITED to be interviewing John Jumper, newly minted Nobel laureate and Google DeepMind director (and team leader of AlphaFold) on stage in London next Tuesday. Tune in to http://
Bloomberg.com to check it out! (Also live on the Terminal!) -
Diffusion Models for Imaging and Vision Tutorial
By
–
Tutorial on Diffusion Models for Imaging and Vision https://
bit.ly/3VCIdCs
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
How to re-rank your snippets in RAG with ColBERT, Rerankers
By
–
𝗛𝗼𝘄 𝘁𝗼 𝗿𝗲-𝗿𝗮𝗻𝗸 𝘆𝗼𝘂𝗿 𝘀𝗻𝗶𝗽𝗽𝗲𝘁𝘀 𝗶𝗻 𝗥𝗔𝗚 ⇒ ColBERT, Rerankers, Cross-Encoders Let’s say you’re doing RAG, and in an effort to improve performance, you try to rerank a few possible source snippets by their relevancy to a query. How can you score
-
Bupa Leverages SAS Analytics for Efficient Healthcare Delivery
By
–
.
@BupaUK Insurance leverages SAS analytics to deliver top-notch, affordable health care while fighting fraud. With real-time insights, Bupa ensures that 3.8M+ UK customers receive fast and efficient care. -
RDT-1B: Largest Diffusion Foundation Model for Robot Manipulation
By
–
Great work from Tsinghua University! RDT-1B, the largest diffusion-based foundation model for robotic manipulation.
RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation https://
rdt-robotics.github.io/rdt-robotics https://
arxiv.org/pdf/2410.07864 -
Evolved Transformer uses 25% less energy than standard models
By
–
The last few rows of Table 1 in https://
arxiv.org/abs/2104.10350 report effect of this on energy usage and CO2e emissions: the more accurate Evolved Transformer (Medium) model uses 25% less energy to train on TPUv2 than does the Transformer (Big) model (& similar story on P100 GPUs) -
Evolved Transformer Achieves Better BLEU Score With 37.6% Fewer Parameters
By
–
See Figure 4 in section 5.3 of So et al (
https://
arxiv.org/abs/1901.11117) In the figure, for example, the second-to-last red dot (Evolved Transformer) achieves a higher BLEU score than the last and largest blue dot configuration (plain vanilla transformer) & has 37.6% fewer parameters.