omg stahp, you're too kind, it's just SLMs, i could go on alllll dayyyy!
MACHINE LEARNING
-
Discussing Small Language Models: Key Insights
By
–
i'm SO HAPPY because i'm discussing SMALL LANGUAGE MODELS https://t.co/3fBreKzKGr
— Rachel Metz (@rachelmetz) 8 août 2024i'm SO HAPPY because i'm discussing SMALL LANGUAGE MODELS
-
Customizing Open-Source Small Language Models: Performance and Getting Started
By
–
What does it take to customize open-source #small language models? How do they #perform? How can you get started? @grg_arnav answered those questions and more during his talk at the @arizeai Observe Conf Check the replay:
-
SIGGRAPH 2024: AI Expands Into Robotics and Autonomous Vehicles
By
–
We had an incredible week at #SIGGRAPH2024. See how SIGGRAPH is leading conversations on computer graphics, simulation and #AI as it expands beyond media and entertainment — into huge industries like robotics, autonomous vehicles and scientific discovery. https://
youtube.com/watch?v=01roO7
9hWNA&list=PLZHnYvH1qtOZ7aKwtMtGamMrF406x-Y6O
… -
Maximizing Llama Energy Efficiency Sweet Spot
By
–
The Sweet Spot – Maximizing Llama Energy Efficiency · Jacques Mattheij https://
bit.ly/4eVq6im
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Math Problems Solved by Recombining Training Data Patterns
By
–
Much of the math problems that seem hard to us can actually be solved by basically sticking together stuff in the training data with minor variations.
-
Language Models Struggle with Multi-Digit Multiplication Tasks
By
–
There are no language models that do multi-digit multiplication well as yet AFAIK, although there is one that can do addition. We discuss that example at length here BTW: https://
answer.ai/posts/2024-07-
25-transformers-as-matchers.html
… -
Meta’s Llama 3 405B Rivals OpenAI GPT-4o Performance
By
–
Thrilled to have the Llama team from @AIatMeta answering questions about their new Llama 3 paper! The newest 405B variant is currently the best performing open source model and rivals OpenAI's GPT-4o across a variety of benchmarks.
-
Qwen2-Math: New Chinese AI Model Surpasses Open Source and Commercial Models
By
–
¡Nuevo Modelo QWEN2-MATH! El laboratorio chino ha hecho público su nuevo modelo Qwen2-Math especializado para matemáticas, y que supera tanto a los modelos open source como a los comerciales. El razonamiento y las matemáticas: el próximo reto a batir por todos los labs!
-
Small Language Models: The Shift from Bigger to Smaller AI
By
–
you know how AI companies have been laser-focused on making bigger and bigger AI models? now they're focused on making them smaller and smaller! (think cheaper, more focused, more efficient.) read my latest on small language models, or SLMs: