AI Digest #178: World’s largest and most powerful generative language model
#178 — October 18, 2021 | View in browser |
AI Digest
Spread the word, build the community, share the knowledge – invite your friends.
this week's favorite
World’s largest and most powerful generative language model
We are excited to introduce the DeepSpeed- and Megatron-powered Megatron-Turing Natural Language Generation model (MT-NLG), the largest and the most powerful monolithic transformer language model trained to date, with 530 billion parameters. It is the result of a research collaboration between Microsoft and NVIDIA to further parallelize and optimize the training of very large AI models.
DeepMind is developing one algorithm to rule them all
Classical algorithms are what have enabled software to eat the world, but the data they work with does not always reflect the real world. Deep learning is what powers some of the most iconic AI applications today, but deep learning models need retraining to be applied in domains they were not originally designed for.
The 2021 machine learning, AI and data landscape
One story has been the maturation of the ecosystem, with market leaders reaching large scale and ramping up their ambitions for global market domination, in particular through increasingly broad product offerings.
Alias-free generative adversarial networks
We observe that despite their hierarchical convolutional nature, the synthesis process of typical generative adversarial networks depends on absolute pixel coordinates in an unhealthy manner. This manifests itself as, e.g., detail appearing to be glued to image coordinates instead of the surfaces of depicted objects. We trace the root cause to careless signal processing that causes aliasing in the generator network.
We present a fast bottom-up method that jointly detects over 100 keypoints on humans or objects, also referred to as human/object pose estimation. We model all keypoints belonging to a human or an object -- the pose -- as a graph and leverage insights from community detection to quantify the independence of keypoints.
newsletters
Older messages
AI Digest #177: The "Frequently bought together" recommendation system
Sunday, October 10, 2021
And more news, tutorials and articles about AI, machine learning, and data science in this week's issue. #177 — October 11, 2021 View in browser AI Digest Spread the word, build the community,
AI Digest #176: The first rule of machine learning: Start without machine learning
Sunday, October 3, 2021
And more news, tutorials and articles about AI, machine learning, and data science in this week's issue. #176 — October 04, 2021 View in browser AI Digest Spread the word, build the community,
AI Digest #175: What is an A/B Test?
Sunday, September 26, 2021
And more news, tutorials and articles about AI, machine learning, and data science in this week's issue. #175 — September 27, 2021 View in browser AI Digest Spread the word, build the community,
AI Digest #174: The definitive guide to Embeddings
Sunday, September 19, 2021
And more news, tutorials and articles about AI, machine learning, and data science in this week's issue. #174 — September 20, 2021 View in browser AI Digest Spread the word, build the community,
AI Digest #173: Data engineering roadmap
Sunday, September 12, 2021
And more news, tutorials and articles about AI, machine learning, and data science in this week's issue. #173 — September 13, 2021 View in browser AI Digest Spread the word, build the community,
You Might Also Like
Architecture Weekly #178 - 6th May 2024
Monday, May 6, 2024
This time, we discussed biases. Biases on the perspective on our technologies, so not seeing their evolutions. We also checked how biases can impact our knowledge, collaboration and eventually also the
WP Weekly 192 - WP Biz - Brands Merged, Woo Cart Popup, Fastest Hosting
Monday, May 6, 2024
Read on Website WP Weekly 192 / WP Biz The 'business of WordPress' is buzzing for sure, be it the acquisition of plugins or the massive Envato ownership change. Also, WordPress content brands
SRE Weekly Issue #423
Monday, May 6, 2024
View on sreweekly.com A message from our sponsor, FireHydrant: FireHydrant is now AI-powered for faster, smarter incidents! Power up your incidents with auto-generated real-time summaries,
⚠️ Avoiding AI Scams on Social Media — An Open Source Google Photos Alternative
Sunday, May 5, 2024
Also: Reviewing the Customizable Drop Mechanical Keyboard, and More! How-To Geek Logo May 5, 2024 📩 Get expert reviews, the hottest deals, how-to's, breaking news, and more delivered directly to
Daily Coding Problem: Problem #1432 [Medium]
Sunday, May 5, 2024
Daily Coding Problem Good morning! Here's your coding interview problem for today. This question was asked by Snapchat. Given the head to a singly linked list, where each node also has a “random”
PD#572 Good Ideas in Computer Science
Sunday, May 5, 2024
Ideas every programmer likes and why Garbage Collection and Object Oriented Programming don't count
RD#454 API Layer & Fetch Functions
Sunday, May 5, 2024
ixing API and UI code quickly leads to messy and unmaintainable code
The Shiny Toy Syndrome & Tiny macOS utility apps I love
Sunday, May 5, 2024
Lex launching its redesign, Raycast shares another monthly update packed with AI updates, prompts should be designed not engineered, and a lot more in this week's issue of Creativerly. Creativerly
Hyundai antes up $1B for AV startup Motional and Elon unplugs the Tesla Supercharger team
Sunday, May 5, 2024
Plus, layoffs come for Luminar, Fisker and Ola View this email online in your browser By Kirsten Korosec Sunday, May 5, 2024 Image Credits: Motional Welcome back to TechCrunch Mobility — your central
C#504 Adventures serializing absolutely everything in C#
Sunday, May 5, 2024
A fantastic journey porting Newtonsoft.Json to System.Text.Json