AI BriefingHuggingFace Blog news
The latest AI and developer news from HuggingFace Blog, with short summaries.
Visit website →Building RAG-Based Data Sovereignty at a French Public Institution
France's Banque des Territoires built a RAG solution that ensures data sovereignty.
Hugging Face Adds Google TPU Support
Google Cloud TPU v5e is now available in Hugging Face's Inference Endpoints and Spaces.
Hugging Face Enhances Dataset Search Features
Hugging Face has introduced dataset search capabilities by modality, size, format, and library.
Gaudi 2 Accelerates Protein Model ProtST
Intel Gaudi 2 accelerators significantly improved the inference and fine-tuning performance of the protein language model ProtST compared to A100.
HuggingFace releases smolagents after taking No. 1 on GAIA
HuggingFace's Code Agent achieved the top rank on the challenging GAIA benchmark, and the related functionality has been spun off into the smolagents library.
Google Unveils Next-Generation Open LLM Gemma 2
Google has unveiled Gemma 2, its next-generation open LLM with significantly improved performance and efficiency.
Patent-specialized embedding model ParaEmbed 2.0 released
XLSCOUT collaborated with Hugging Face to launch ParaEmbed 2.0, optimized for patent and intellectual property (IP) analysis.
Key Factors of Data Quality That Determine AI Performance
Building high-quality AI models requires data relevance, comprehensiveness, timeliness, and bias mitigation.
Florence-2, DocVQA Fine-Tuning Guide
Introduces how to fine-tune Microsoft's small vision-language model Florence-2 for DocVQA tasks and the resulting performance.
Hugging Face Announces Open-Source Dataset Achievements
Hugging Face unveiled the dataset-building achievements and future plans of the collaborative project 'Data Is Better Together'.
Code Generation Benchmark BigCodeBench Released
BigCodeBench, a new LLM code generation benchmark for evaluating practical programming ability, has been released.
Analyzing the Precision Handling Differences Between DeepSpeed and FSDP
This analyzes how differences in the internal precision handling methods of DeepSpeed and PyTorch FSDP affect model training outcomes.
HuggingFace Releases RLOO Trainer
HuggingFace has introduced RLOO Trainer, a new RLHF algorithm that is more memory-efficient and faster than PPO, into TRL.
Diffusers Adds Support for Stable Diffusion 3
Hugging Face's Diffusers library now officially supports the Stable Diffusion 3 (SD3) model.
Hugging Face Launches Embedding Container for SageMaker
Hugging Face has officially launched a dedicated container for efficiently deploying embedding models on Amazon SageMaker.
Artificial Analysis unveils image generation model leaderboard
A leaderboard comparing the performance of image generation models based on over 45,000 human preferences has launched.
LLM-Based 3D NPC Interaction Platform Unveiled
NPC-Playground, an open-source project that implements intelligent NPCs in 3D environments using LLMs' function calling capabilities, has been unveiled.
Intel Gaudi Boosts Inference Speed with Assisted Generation Support
Intel Gaudi now supports Speculative Sampling-based Assisted Generation, boosting LLM inference speed by up to 2x.
HuggingFace Responds to Spaces Security Incident
HuggingFace detected a security incident on the Spaces platform and implemented token invalidation and enhanced security measures.
Hugging Face Releases TGI Optimization Tool
It introduces a benchmarking tool that analyzes TGI's performance metrics and how to use it to optimize LLM deployment.