AI BriefingHuggingFace Blog news
The latest AI and developer news from HuggingFace Blog, with short summaries.
Visit website →Hugging Face Demos Integrated into arXiv
A feature has been integrated that allows machine learning demos to be run instantly from arXiv paper pages via Hugging Face Spaces.
HuggingFace Revamps Inference API
HuggingFace is ending the paid tier of the Inference API and introducing a new pricing structure including Inference Endpoints.
Contrastive Search Introduced to Transformers
Contrastive Search, which solves the text repetition problem, has been introduced to the Transformers library.
Dreambooth Training Optimization Guide Based on Diffusers
HuggingFace has released optimal settings for preventing overfitting and improving quality when training Dreambooth on Stable Diffusion using Diffusers.
Guide to Fine-Tuning Whisper for Multilingual ASR
This guide explains how to fine-tune the Whisper model on multilingual automatic speech recognition (ASR) datasets using Hugging Face Transformers.
Hugging Face Integrates Intel OpenVINO
Intel OpenVINO has been integrated into Hugging Face's Optimum Intel, making it easier to run model inference and quantization on Intel processors.
🤗 Evaluate Adds Bias Evaluation Feature
Hugging Face has added bias evaluation metrics to the 🤗 Evaluate library that can measure toxicity, polarity, and hurtfulness in language models.
MTEB: Text Embedding Performance Evaluation Benchmark Released
MTEB, a large-scale text embedding benchmark supporting 8 tasks and 112 languages, has been released.
Guide to Using Hugging Face Inference Endpoints
This introduces how to use Inference Endpoints, which allow easy and secure deployment of models from the Hugging Face Hub to managed infrastructure.
Diffusers adds JAX/Flax support
Hugging Face's Diffusers library now supports Flax, enabling fast inference of Stable Diffusion on TPU environments.
Technical Analysis of BLOOM's 50x Inference Performance Boost
HuggingFace explains the technical process of optimizing the BLOOM model's inference server, improving latency by 5x and throughput by 50x.
Hugging Face Adds DOI Support for Models and Datasets
Hugging Face has introduced DOI generation for permanent citation and reproducibility of models and datasets.
Japanese-specialized Stable Diffusion released
rinna has released 'Japanese Stable Diffusion,' which reflects Japanese prompts and Japanese culture.
AutoTrain adds support for image classification
Hugging Face's no-code ML tool AutoTrain now supports image classification.
How Accelerate Loads Large Models
It explains how to leverage PyTorch's meta device to overcome memory constraints and efficiently load large language models.
Efficient Few-Shot Learning Framework SetFit Released
SetFit, a few-shot learning framework that achieves high performance with little data and no prompts, has been released.
BLOOM Inference: DeepSpeed vs Accelerate
Compares the optimal hardware configuration and benchmarks for maximizing inference performance of the 176B-parameter BLOOM model using DeepSpeed and Accelerate.
Diffusers Library New Feature Update
Hugging Face's Diffusers library has released version 0.3, adding key features such as Image-to-Image and Textual Inversion.
Decision Transformer Learning Guide
Introduces the concept and training method of Decision Transformer, which solves reinforcement learning as sequence modeling.
OpenRAIL Proposed as an AI-Specific License
It proposes the OpenRAIL license, which addresses the limitations of existing open-source licenses to govern the responsible use of AI models.