✷ The archive · 11 labs · 871 dispatches
The labs
No more opening a dozen official blogs every morning. First-hand releases from OpenAI, Anthropic, DeepMind and the rest, each with its substance pulled out.
Large-scale Near-deduplication Behind BigCode
Hugging Face details the use of MinHash and Locality Sensitive Hashing (LSH) to perform large-scale near-deduplication for the BigCode project, demonstrating that removing near-duplicate data improves model performance and training efficiency.
RWKV Architecture Integration into Hugging Face Transformers
Hugging Face announced the integration of the RWKV RNN‑Transformer hybrid model into the Transformers library, enabling open‑source, long‑context language models that combine RNN efficiency with transformer performance.
Running Vicuna 13B on AMD GPUs with ROCm
Hugging Face provides a technical guide on deploying the Vicuna 13B open-source chatbot on a single AMD GPU using ROCm and GPTQ 4-bit quantization to reduce memory requirements from 28GB to 7.52GB.
Hugging Face Selected for CNIL Enhanced Support Program
Hugging Face has been selected by the French Data Protection Authority (CNIL) for its Enhanced Support program to improve the implementation of data protection and GDPR compliance in AI development.
Hugging Face Assisted Generation for Low-Latency Text Generation
Hugging Face introduced Assisted Generation, a decoding method that uses a smaller assistant model to predict candidate tokens which are then validated by a larger model, reducing latency by up to 10x in some hardware configurations.
Creating a Coding Assistant with StarCoder – Summary
This article describes how the Hugging Face team turned the 16 B‑parameter StarCoder code‑generation model into a conversational coding assistant (StarChat‑α). By adding special chat tokens, preparing a filtered Open‑Assistant dialogue dataset, and fine‑tuning with DeepSpeed ZeRO‑3, they trained a model that can understand user prompts and generate runnable code (e.g., plots, maps, visualisations). The post also covers token‑izer configuration, loss‑masking for user turns, training setup, evaluation (both benchmark and LLM‑based), limitations, and future directions. The resulting model is publicly available on the Hub.
A Dive into Text-to-Video Models
Hugging Face provides a technical overview of the evolution of text-to-video generative models, detailing the transition from GANs to Transformers and Diffusion architectures while highlighting the unique challenges of temporal consistency and data scarcity.
StarCoder Release Notes
Hugging Face and ServiceNow's BigCode collaboration released StarCoder and StarCoderBase, 15B parameter Code LLMs trained on permissively licensed data that outperform several open and closed models on programming benchmarks.
Hugging Face Unity API Installation and Usage Guide
The Hugging Face Unity API allows developers to integrate Hugging Face Inference API models into Unity projects via a dedicated package and API wizard.
Training Language Models with Hugging Face Transformers, TensorFlow, and TPUs
Hugging Face provides a scalable end-to-end guide for training masked language models from scratch using TensorFlow and TPU pods, leveraging XLA compatibility for high-performance compute.
Databricks and Hugging Face Integrate Apache Spark for Faster LLM Training
Databricks has introduced first-class Apache Spark support in Hugging Face Datasets via the Dataset.from_spark() function, reducing data loading times by up to 40% for large-scale model training and tuning.
Running DeepFloyd IF on Free‑Tier Google Colab with Diffusers
Hugging Face shows how to run the open‑source DeepFloyd IF text‑to‑image model on a free Google Colab notebook by using 8‑bit quantization, modular pipeline loading, and Diffusers‑integrated optimizations.
Hugging Face Launches Chinese Language Blog to Support Global AI Community
Hugging Face has launched a dedicated Chinese language blog (hf.co/blog/zh) to provide translated technical resources and foster deeper collaboration with the Chinese AI community.
Hosting Unity Games on Hugging Face Spaces
Hugging Face Spaces can host playable Unity games by utilizing the Static HTML template and WebGL build targets.
Accelerating Hugging Face Transformers with AWS Inferentia2
Hugging Face and AWS have optimized Transformers for AWS Inferentia2, a purpose-built inference accelerator that significantly reduces latency and increases throughput for large-scale models.
Graph Classification with Transformers
Hugging Face demonstrates how to perform graph classification using the Graphormer model within the Transformers library, covering data formatting, preprocessing, and fine-tuning.
Creating Privacy Preserving AI with Substra
Hugging Face and Substra highlight how federated learning enables the training of AI models across decentralized data sources to maintain privacy and security, particularly in sensitive domains like healthcare.
Snorkel AI and Hugging Face Integration for Enterprise Foundation Models
Snorkel AI has partnered with Hugging Face to integrate Hugging Face Inference Endpoints into Snorkel Flow, allowing enterprises to easily access and adapt over 150,000 open-source foundation models.
StackLLaMA: Training LLaMA with RLHF for Stack Exchange
Hugging Face introduces StackLLaMA, a model trained using Supervised Fine-tuning, Reward Modeling, and Reinforcement Learning from Human Feedback (RLHF) to answer Stack Exchange questions.
Hugging Face Ethics and Society Newsletter #3: Ethical Openness Initiative
Hugging Face announced new ethical openness measures—including six ethics tags, a flagging system, and audience‑guiding metadata—to make open‑source ML safer and more inclusive.
Accelerating Stable Diffusion Inference on Intel Sapphire Rapids CPUs
Hugging Face shows how to cut Stable Diffusion image generation from 32 seconds to about 5 seconds on Intel Sapphire Rapids CPUs using Optimum Intel, OpenVINO, system‑level tweaks, IPEX BF16, and a faster scheduler.
BLOOMZ Inference on Habana Gaudi2 Accelerator
Hugging Face demonstrates that the Habana Gaudi2 accelerator achieves faster inference for the BLOOMZ 176B model than the Nvidia A100 80GB, leveraging the Optimum Habana library and DeepSpeed-inference.
Federated Learning with Hugging Face and Flower
Hugging Face demonstrates how to use the Flower framework to perform federated learning on a pre-trained distilBERT model for sentiment analysis on the IMDB dataset.
Training ControlNet with Hugging Face Diffusers
Hugging Face provides a comprehensive guide and training script via the diffusers library to enable users to train custom ControlNet models for Stable Diffusion, demonstrated through the creation of an 'Uncanny Faces' pose model.
Hugging Face Hub Improved Jupyter Notebook Support
Hugging Face has introduced human-readable rendering for Jupyter notebooks hosted on the Hub, improving reproducibility and accessibility for machine learning practitioners.
Multivariate Probabilistic Time Series Forecasting with Informer
Hugging Face has integrated the Informer model into the Transformers library to enable efficient multivariate probabilistic time series forecasting with reduced computational and memory complexity.
Fine-tuning 20B LLMs with RLHF on a 24GB Consumer GPU
Hugging Face released an integration of TRL and PEFT that enables reinforcement learning fine‑tuning of 20‑billion‑parameter language models on a single 24 GB GPU using 8‑bit quantization and low‑rank adapters.
Kakao Brain ViT and ALIGN Models Release with COYO 700M Dataset
Kakao Brain and Hugging Face released open‑source ViT and ALIGN visual‑language models trained on the new 700 M image‑text COYO dataset, providing the first publicly available ALIGN model and ViT models with reproducible training data.
ControlNet in Diffusers
Hugging Face has integrated ControlNet into the Diffusers library, enabling precise spatial control over Stable Diffusion image generation using conditionings like Canny edges, depth maps, and human poses.
Using Machine Learning for Disaster Response: The afetharita Project
Hugging Face describes how volunteers used the Hugging Face ecosystem to rapidly deploy ML models for OCR, NER, and remote sensing to aid survivors of the February 2023 Turkey earthquakes.
Hugging Face Diffusers Ethical Guidelines
Hugging Face has introduced an ethical framework for the Diffusers library to guide technical decisions and community contributions while mitigating the potential societal risks of diffusion models.
Hugging Face Expert Acceleration Program: Witty Works Case Study
Witty Works utilized the Hugging Face Expert Acceleration Program and SetFit to build a context-dependent inclusive language classifier with high accuracy using minimal labeled data.
Swift Diffusers for Mac 1.1 Release
Hugging Face has released Diffusers for Mac version 1.1, a native open-source app that leverages Core ML to accelerate Stable Diffusion on Apple Silicon, offering up to 2x faster generation on certain hardware configurations.
Red-Teaming Large Language Models
Hugging Face outlines the critical role of red-teaming in identifying LLM vulnerabilities to prevent harmful outputs, emphasizing the need for collaborative, adaptive evaluation methods.
Fetch AI Infrastructure Migration: Consolidating Tools with Hugging Face and AWS
Fetch reduced development time by 30% and processing latency by 50% by migrating from a third-party AI 'black box' to an in-house ML pipeline powered by Hugging Face and AWS.
Hugging Face and AWS Strategic Partnership for AI Accessibility
Hugging Face and AWS have expanded their strategic partnership to democratize generative AI by integrating Hugging Face models with AWS infrastructure and purpose-built ML accelerators.
Hugging Face Inference Endpoints Case Study
Hugging Face is migrating its CPU-based ML models from AWS ECS/Fargate to Hugging Face Inference Endpoints to reduce operational complexity and improve latency.
BLIP-2: Zero-Shot Image-to-Text Generation
BLIP-2 is a visual-language model from Salesforce Research that uses a Querying Transformer (Q-Former) to bridge frozen image encoders and frozen large language models for efficient zero-shot image-to-text generation.
🤗 PEFT library release enables parameter-efficient fine-tuning of billion‑scale models
Hugging Face announced the 🤗 PEFT library, which lets users fine‑tune billion‑parameter models on low‑resource hardware by training only a tiny fraction of parameters.
SpeechT5: Unified-Modal Encoder-Decoder for Speech Processing
Hugging Face has integrated SpeechT5, a unified Transformer-based model capable of text-to-speech, speech-to-speech, and speech-to-text tasks through a shared encoder-decoder backbone.
Hugging Face AI vs. AI: Multi-Agent Reinforcement Learning Competition System
Hugging Face has introduced AI vs. AI, an open-source system for ranking deep reinforcement learning models through continuous multi-agent competitions using an ELO rating system.
Hugging Face AI for Game Development: Generating Stories
Hugging Face outlines a workflow for using Large Language Models (LLMs) like ChatGPT to generate game stories and content, while highlighting critical limitations regarding originality, legal risks, and long-term coherence.
Accelerating PyTorch Transformers with Intel Sapphire Rapids
Hugging Face demonstrates that combining Intel Sapphire Rapids CPUs with the Optimum Intel library can accelerate PyTorch transformer inference by up to 3x compared to previous Xeon generations.
A Dive into Vision-Language Models
Hugging Face provides a technical overview of vision-language models, detailing five primary pre-training strategies and their integration into the Transformers library for tasks like VQA and image segmentation.
The State of Computer Vision at Hugging Face
Hugging Face has expanded its ecosystem to support 8 core computer vision tasks, over 3,000 models, and 100+ datasets, integrating both Transformer and convolutional architectures.
Hugging Face 2D Asset Generation for Game Development
Hugging Face demonstrates how to integrate Stable Diffusion's Image2Image capability into a collaborative 2D game asset workflow to create production-ready icons efficiently.
Using LoRA for Efficient Stable Diffusion Fine-Tuning
Hugging Face has integrated Low-Rank Adaptation (LoRA) into the diffusers library, enabling faster Stable Diffusion fine-tuning with significantly lower VRAM requirements and model weights as small as 3 MB.
Hugging Face Optimum and ONNX Runtime Training Integration
Hugging Face and Microsoft have integrated ONNX Runtime into the Optimum library to reduce training times for transformer-based models by 35% or more.
What Makes a Dialog Agent Useful? Technical Analysis
Hugging Face analyzes the key techniques—including Instruction Fine-Tuning (IFT), Supervised Fine-Tuning (SFT), Reinforcement Learning from Human Feedback (RLHF), and Chain-of-Thought (CoT)—that transform base language models into useful conversational agents.
Hugging Face AI for Game Development: 3D Asset Generation
Hugging Face explores the current state of text-to-3D AI, concluding that while tools like DreamFusion and Point-E exist, they are not yet practically applicable for rapid game development workflows.