Artificial Intelligence #31
Andriy Burkov
PhD in AI, author of ?? The Hundred-Page Language Models Book and ?? The Hundred-Page Machine Learning Book, ML at TalentNeuron
Hey, in this issue: machine learning of sets, video lectures on deep learning for computer vision, robot playing badminton, Bayesian deep learning, a new framework for distributed reinforcement learning, and more.
The sponsor of this issue is Colibri.ai.
- Machine learning of sets
- Deep learning for computer vision (video lectures)
- Robot playing badminton (video)
- A minimal PyTorch implementation of the OpenAI GPT training
- [Sponsored] Colibri.ai: Automated meeting notes
Colibri records online meetings, transcribes them in real-time and generates concise searchable meeting notes. Works with Zoom, Google Meet, Jitsi, etc. Available white-label, and on-prem, with custom language models and NLP filtering utilities. Get early access
- PyTorch Lightning is a Keras of PyTorch (step-by-step, tutorial, getting started)
- On Bayesian deep learning (video)
- How smart is BERT? Evaluating the language model's commonsense knowledge
- Machine learning helps map global ocean communities
- Coronavirus accelerates the adoption of AI in health care
- Acme: A new framework for distributed reinforcement learning
If you build an AI or data product or service, you are invited to sponsor one of the next issues. Feel free to contact True Positive Inc. for more details on sponsorships.
Enjoy the newsletter? Help us make it bigger and better by sharing with your colleagues and friends.
Have a nice weekend! See you next week. — Andriy
Education Professional at Education Week
4 年spiritual healing heals mind body and soul https://www.dreamfulllife.com/2020/09/spiritual-healing.html
Sales Leadership: Better Business Thru Technology
4 年Great collection. The short BERT article (and associated paper) was interesting -- sort of along the lines of "does the BERT language framework use common sense to answer questions?" Apparently the answer is "yes"! Which is weird because then we have to conclude that BERT is "smarter than we knew", because it took some serious researchers running experiments to show that BERT was in fact using a little bit of common sense. Common sense wasn't just deliberately coded. Can we call this emergent behaviour? Now I'm curious if such common sense might be part of other machine learning systems, for example for field service and maintenance. Beyond just pattern recognition, for a field service AI system to have some sense of real world semantics would be powerful. Am I reading too much into this?
ICT and Project Management Professional | Scrum Master | Business Analyst and Strategic Management | ISO27001A | Official Network University and Senior Official of International Association of Project Managers (IAPM)
4 年Thank you Andriy Burkov. Esp #31 NLP for Sentiment Analysis is my current research. Keep inspiring us chaps!
Building @STEM Spectrum | Data Science | Business Automation | LinkedIn Marketing | FinTech | AI ML | Cosmology Enthusiast | Networking & Learning
4 年Nice article