Scott McCarty’s Post

Nerlnet's Wiki is updated with instructions of how to use the Nerlplanner GUI to generate DC files. https://lnkd.in/dTDTqMx2 #nerlplanner #nerlnet #distributed_ml #distributed_systems #erlang #nn #cnn #autoencoder

https://lnkd.in/gkuy6vH9 Fascinating that we are seeing a wave of LLLms that are simultaneously more accurate and more compute efficient. Already like Claude Sonnet and 3.5 looks like a strong advance

More details about the mighty phi-3 family of SLMs. Such a key development as GenAI Dev becomes more about always optimizing and orchestrating with LLMs routing to LLMs and SLMs continuously to balance complexity, latency, cost and more! https://lnkd.in/gGQYqNKt

Best Resources to Learn & Understand Evaluating LLMs via #TowardsAI → https://bit.ly/3WAPbs2

Best Resources to Learn & Understand Evaluating LLMs via #TowardsAI → https://bit.ly/3WAPbs2

Our Forbes article on the limitations of LLMs, and how Tensor Networks can help, is finally out!

Great Insight - Mixture-of-Experts (MoE): The Birth and Rise of Conditional Computation “As the training of giant dense models hits the boundary on the availability and capability of the hardware resources today, Mixture-of-Experts (MoE) models have become one of the most promising model architectures due to their significant training cost reduction compared to quality equivalent dense models.” - from [12] #genai #llm #chatgpt URL : https://lnkd.in/gE_CemwR

The skepticism continues over whether Reflection 70B offers genuine innovation or is simply an overhyped iteration. 🤔 This underscores the critical need for transparency, reproducibility, and caution before rushing to claim AI superiority. In such a rapidly evolving field, verifiable results are more important than ever. 🔍✅ What’s your take on the Reflection 70B debate? #AI #MachineLearning #Transparency #AIDevelopment #LLM https://lnkd.in/gpJFebH8

my focus is on algorithm compression not more compute power. biggest model I aim for is 7b

AGI achieved by scaling compute and talk of ASI. https://lnkd.in/gJykUtBZ