AI BriefingHuggingFace Blog news
The latest AI and developer news from HuggingFace Blog, with short summaries.
Visit website →Deploying AI Comic Factory with the Hugging Face Inference API
This guide explains how to deploy the popular AI application AI Comic Factory to your own Space using Hugging Face's Inference API.
TRL Supports Stable Diffusion RLHF with DDPO
HuggingFace's TRL library now supports DDPO, enabling RLHF fine-tuning of Stable Diffusion models.
SageMaker Llama 2 Deployment Benchmark
Hugging Face has released benchmark results for over 60 configurations to optimally deploy Llama 2 on Amazon SageMaker.
Hugging Face Launches Inference API for PRO
Hugging Face has launched 'Inference for PRO,' offering PRO subscribers a dedicated API for high-performance models with higher usage limits.
Rocket Money scales ML with Hugging Face
Rocket Money successfully deployed a BERT model with more than 4,000 classes into production using the Hugging Face Inference API.
Hugging Face unveils Object Detection Leaderboard
Hugging Face released a leaderboard comparing the performance of object detection models and explained the key evaluation metrics.
Overview of 3D Gaussian Splatting
Explains the principles of 3D Gaussian Splatting, which renders photorealistic 3D scenes in real time.
Guide to Optimizing LLM Production Deployment
Covers quantization, Flash Attention, and architectural innovations for efficient production deployment of LLMs.
Guide to Efficient Fine-tuning of Llama 2 70B
It covers how to fine-tune the Llama 2 70B model using PyTorch FSDP while solving memory and checkpoint saving issues.
High-Compression-Based Würstchen Released
Würstchen, a new diffusion model that dramatically reduces training and inference costs through a high spatial compression ratio of 42x, has been released.
🤗 Transformers Supported Quantization Methods Guide
Compares the characteristics of the bitsandbytes and auto-gptq quantization methods natively supported by the 🤗 Transformers library.
Hugging Face Unveils SafeCoder for Enterprise
Hugging Face has unveiled SafeCoder, an enterprise code assistant built on StarCoder that offers transparency and customization.
T2I-Adapter Support and Efficiency Improvements for SDXL
T2I-Adapter, which is much lighter and faster than ControlNet while maintaining SDXL's control performance, has been introduced into diffusers.
TII Releases Falcon 180B, a 180B-Scale Model
TII has released Falcon 180B, a large-scale open-source LLM with 180 billion parameters.
Fetch Cuts ML Latency by 50% with SageMaker
Fetch optimized its ML pipeline using SageMaker and Hugging Face, reducing latency by 50% and increasing accuracy by 200%.
AudioLDM 2 Inference Speed Improved 10x
AudioLDM 2's inference speed has been improved by more than 10x through Hugging Face Diffusers.
Hugging Face Discontinues Git Password Authentication
To strengthen security, Hugging Face is discontinuing password authentication for Git and requiring the use of tokens or SSH keys instead.
Coding-specialized model Code Llama released
Code Llama, an open-source model family based on Llama 2 and optimized for coding tasks, has been released.
Transformers Supports GPTQ
Hugging Face integrates AutoGPTQ into Transformers to support 2-8 bit quantization of LLMs.
HuggingFace Releases IDEFICS
HuggingFace has released IDEFICS, an open-source reproduction of Flamingo, a multimodal model.