Totally missed that and that sounds awesome. Genuine question, how do you leverage unified memory in PyTorch though? Was discussing this with a user who was using Blackwell GH200 the other day here:
MACHINE LEARNING
-
Edge AI Tracks Dozens Vehicles Real-Time Traffic Video
By
–
🚗 Real-time vehicle tracking with Edge AI! 🚦
— Axelera AI (@AxeleraAI) 13 janvier 2025
Axelera AI’s hardware tracks dozens of cars in traffic video – all in real time, processed locally.
Smarter, faster, and more efficient AI for next gen smart cities & transport.https://t.co/4Anjxke494#EdgeAI #ComputerVision pic.twitter.com/QuWvGobdXVReal-time vehicle tracking with Edge AI! Axelera AI’s hardware tracks dozens of cars in traffic video – all in real time, processed locally. Smarter, faster, and more efficient AI for next gen smart cities & transport. https://
axelera.ai #EdgeAI #ComputerVision -
AI and Algorithms Reshape Marketing Landscape and Decision-Making
By
–
Thank you, @ThuliumCo
, for sharing this! The rise of algorithms and AI is truly reshaping the marketing landscape at a pace that challenges all of us to think beyond the present. While machines may take on a larger role in decision-making and even communication, I firmly -
LLMs for Health Equity: Beyond Harms to Opportunities
By
–
Our article on using LLMs to improve health equity is out in @NEJM_AI! 85% of equity-related LLM papers focus on *harms*. But equally vital are the equity-related *opportunities* LLMs create: detecting bias, extracting structured data, and improving access to health info.
-
Diffusion Meets Flow Matching: Advanced Deep Learning Techniques
By
–
Diffusion Meets Flow Matching https://
bit.ly/3VOaNQy
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Mistral AI announces Codestral 25.01 with 256k context
By
–
BREAKING : Mistral AI is preparing to release Codestral 25.01 with 256k context length "Code at the speed of Tab. Available today in Continue dot dev and soon on other leading AI code assistants." h/t @ai_for_success
-
Can Artificial Intelligence’s Hallucination Problem Be Solved?
By
–
Artificial Intelligence sometimes invents things – can this problem be fixed?
#AI #AIio #BigData #ML #NLU #Futureofwork @TunstallAsc @StrategyFintech @TamaraMcCleary @TerenceLeungSF @psb_dc @thomaspower @vinod1975 @ylecun http://
ow.ly/AU0x30sJars -
Loading Pretrained Weights and Selecting Foundation Models
By
–
And in addition, as you will see later in chapter 5, it's necessary to load pretrained weights.
Now there are many open-weight models available, but my reason for this one was also that it's a) the basis for everything that came after (other architectures can be covered in -
Book Update: Future Plans for RNN, LSTM, and GAN Content Revision
By
–
Thanks for your interest in this. I hope to find the time some day, but it won't be very soon I'm afraid. I agree with your suggestions regarding RNN, LSTM, GANs (would replace them with more LLM and diffusion model contents). Unfortunately, the book was already at the max page
-
Next-Token Prediction Self-Supervised Learning and Uncertainty
By
–
This particular view is a decade out of date. You know that next-token prediction is mostly self-supervised, and as such the problem-solution pairs are "automatically" generated? Doing the same to train "I don't know next token" is a minor reframe, much research exists.