AI BriefingHuggingFace Blog news
The latest AI and developer news from HuggingFace Blog, with short summaries.
Visit website →HuggingFace releases smolagents, a code-based agent library
HuggingFace has released smolagents, a simple agent library in which LLMs write code directly to use tools.
PyTorch GPU Memory Visualization Guide
Explains how to visualize and analyze GPU memory usage using PyTorch's memory snapshot feature.
NVIDIA Releases LogitsProcessorZoo for LLM Generation Control
NVIDIA has released LogitsProcessorZoo, a Hugging Face-compatible library for controlling LLM generation.
Big Bench Audio Benchmark Released
A new dataset called Big Bench Audio, designed to evaluate the reasoning capabilities of audio language models, has been released.
ModernBERT Released: The Next-Generation BERT Replacement
ModernBERT, a next-generation BERT replacement model with an 8,192-token long context and improved performance, has been released.
IBM Unveils Hybrid Mamba2 Model
IBM and researchers from major universities have unveiled Bamba-9B, a hybrid Mamba2 model that maximizes inference efficiency.
GCP 5th Gen Xeon Shows Potential for CPU-Based AI Agents
GCP's 5th generation Xeon (C4) instances demonstrated overwhelming efficiency in text embedding and generation performance compared to the previous generation.
TII Unveils Falcon 3 Open Model Family
TII has released the Falcon 3 open model family, featuring under-10B-parameter models with enhanced math, coding, and reasoning capabilities.
Hugging Face Unveils Synthetic Data Generator Based on Natural Language
Synthetic Data Generator, a no-code tool that lets you create custom datasets using only natural language descriptions, has been released.
Hugging Face unveils LeMaterial
Hugging Face and Entalpic have unveiled LeMaterial, an open source project to accelerate materials research, along with the unified dataset LeMat-Bulk.
Open Text-to-Image Preference Dataset Released
The Hugging Face community has released an open-source preference dataset for training text-to-image generation models.
Amazon Bedrock Now Supports 83 Hugging Face Models
Through the Amazon Bedrock Marketplace, 83 open Hugging Face models are now available for use.
Google releases next-generation VLM PaliGemma 2
Google has unveiled PaliGemma 2, a new vision language model (VLM) built on Gemma 2.
AraGen released for Arabic LLM evaluation
HuggingFace has released AraGen, an Arabic LLM benchmark based on the 3C3H metric that evaluates factuality and helpfulness simultaneously.
Optimizing Small Model Performance Using LLMs
This introduces CFM's case of using LLM-based data labeling to boost small model performance while cutting operational costs by up to 80x.
EU AI Act: A Guide for Open Source Developers
This explains the key provisions of the EU AI Act and what open source developers need to comply with, along with the relevant tools.
HuggingFace unveils 2B compact VLM 'SmolVLM'
HuggingFace has released SmolVLM, a 2B-scale open-source vision language model that maximizes memory efficiency.
Hugging Face Overhauls Large Model Transfer Architecture
Hugging Face is introducing a new CAS-based upload/download architecture for efficient transfer of large models and datasets.
Analyzing the Design Principles of RoPE
This article analyzes step by step the design principles and necessity of RoPE, a core technology in modern LLMs.
Japanese LLM Open Leaderboard released
Hugging Face and LLM-jp collaborated to release an open leaderboard evaluating the performance of Japanese LLMs.