Researchers have developed two tools, Argilla and Hugging Face Spaces, that enable communities to collaborate on building high-quality datasets. These...
Researchers have developed a text-generation pipeline that runs on Intel's Gaudi 2 AI accelerator. The pipeline is based on the Hugging Face Transform...
Hugging Face has released two new tools for natural language processing (NLP): StarCoder2 and The Stack v2. These tools are designed to improve the ef...
Hugging Face has released TTS Arena, a benchmarking platform for evaluating text-to-speech (TTS) models. The platform allows researchers and developer...
AI watermarking is a technique used to embed unique identifiers into machine learning models, making it possible to track their ownership and usage. T...
Hugging Face has announced a new technique for fine-tuning its pre-trained models, called Gemma. This method is based on the PEFT (Progressive Embeddi...
Hugging Face has introduced a new leaderboard called the Red-Teaming Resistance Leaderboard, developed by HaizLab. The leaderboard is designed to eval...
Matryoshka embedding models are a type of neural network architecture designed for natural language processing tasks. They consist of multiple layers,...
Google has released an open large language model (LLM) called Gemma. It is based on the T5 architecture and can perform a range of tasks, including an...
Hugging Face has introduced the Open Ko-LLM Leaderboard, a benchmarking platform for evaluating large language models in the Korean language. The lead...
The PEFT (PyTorch Empirical Floating-Point Training) project has introduced new merging methods. These methods allow for more efficient training of la...
Researchers have created an open-source framework for generating synthetic data. This can help reduce the costs and environmental impact of collecting...