AI
2023, year of open LLMs
The past year has seen a significant shift towards open Large Language Models (LLMs), with many organizations releasing their models and datasets for ...
AI
AI
Increasing accuracy of pediatric visit notes
A healthcare company called Summer Health is using OpenAI's technology to improve the accuracy of notes taken during pediatric doctor visits. This inv...
AI
Practices for Governing Agentic AI Systems
OpenAI has published a set of practices for governing agentic artificial intelligence systems. These systems have the ability to take actions on their...
AI
Superalignment Fast Grants
OpenAI is offering $10 million in grants for research focused on ensuring the safety and alignment of advanced artificial intelligence. The funding ai...
AI
Weak-to-strong generalization
Researchers at OpenAI are exploring a new approach to training AI models called 'weak-to-strong generalization.' This involves using weak or incomplet...
AI
Partnership with Axel Springer to deepen beneficial use of AI in journalism
Axel Springer, a global publishing house, has partnered with OpenAI to integrate artificial intelligence (AI) technologies into its journalism. This p...
AI
Welcome Mixtral - a SOTA Mixture of Experts on Hugging Face
Hugging Face has introduced Mixtral, a state-of-the-art mixture of experts model. This model combines the strengths of different models to achieve bet...
AI
Mixture of Experts Explained
Mixture of Experts (MoE) is a deep learning technique that combines multiple models to make predictions. It was originally proposed in the 1990s but h...
AI
SetFitABSA: Few-Shot Aspect Based Sentiment Analysis using SetFit
SetFitABSA is a few-shot learning approach for aspect-based sentiment analysis. It uses the SetFit framework to adapt pre-trained language models to s...
AI
AMD + 🤗: Large Language Models Out-of-the-Box Acceleration with AMD GPU
AMD and Hugging Face have collaborated to accelerate large language models on AMD GPUs. The partnership allows users to run these models out-of-the-bo...
AI
Optimum-NVIDIA Unlocking blazingly fast LLM inference in just 1 line of code
Researchers have developed a method called Optimum-NVIDIA that enables fast language model inference using NVIDIA GPUs. This is achieved by optimizing...
AI
Goodbye cold boot - how we made LoRA Inference 300% faster
Researchers have developed a new method called LoRA (Low-Rank Adaptation) that significantly speeds up the inference process in machine learning model...