✷ The archive · 11 labs · 3,433 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.
🤗 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.
Anthropic Partners with Google Cloud for AI Infrastructure
Anthropic has selected Google Cloud as its cloud provider to leverage GPU and TPU clusters for training, scaling, and deploying its AI systems, including the Claude assistant.
ChatGPT Plus launch
OpenAI announced ChatGPT Plus, a $20/month subscription that guarantees access during peak times, faster responses, and priority feature rollouts, expanding its availability worldwide.
OpenAI AI Text Classifier Release (January 2023)
OpenAI released a public AI text classifier in January 2023 that flags likely AI‑generated English text, but it is unreliable—identifying only 26% of AI text and mislabeling 9% of human text.
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.
OpenAI and Microsoft Extend Strategic Partnership
OpenAI and Microsoft have extended their multi-year, multi-billion dollar partnership to accelerate independent AI research and expand the deployment of AI models via Azure.
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.
Mask2Former and OneFormer: Universal Image Segmentation Models in 🤗 Transformers
Hugging Face released Mask2Former and OneFormer in the Transformers library, providing universal architectures that handle instance, semantic, and panoptic segmentation with a single model.
PaddlePaddle Integration with Hugging Face Hub
Hugging Face has partnered with PaddlePaddle to integrate its deep learning platform and libraries, starting with PaddleNLP, into the Hugging Face Hub for improved accessibility and sharing.
Image Similarity with Hugging Face Datasets and Transformers
Hugging Face demonstrates how to build an image similarity system using the Transformers and Datasets libraries by computing dense vector embeddings and measuring cosine similarity.
Forecasting Misuse of Language Models for Disinformation Campaigns
OpenAI, Georgetown University, and the Stanford Internet Observatory released a report analyzing how large language models could lower the cost and increase the scale of disinformation campaigns, proposing a four-stage framework for mitigation.
AI for Game Development: Using LLMs for Game Design
Hugging Face demonstrates how Large Language Models like ChatGPT can be used as brainstorming and acceleration tools for game design, specifically for defining core features of a farming game.
Anthropic Research: Superposition, Memorization, and Double Descent
Anthropic researchers investigate the relationship between superposition, memorization, and overfitting in simple neural networks, suggesting that superposition allows models to store more features than they have neurons during memorization.
Yabble GPT-3 Integration for Customer Feedback Analysis
Yabble integrated OpenAI's GPT-3 to reduce the time required to analyze complex customer feedback data sets from weeks to minutes.
Waymark Fine-tuning GPT-3 for Video Script Creation
Waymark integrated fine-tuned GPT-3 models to automate the creation of tailored, high-quality video commercial scripts, reducing customer editing time and scaling video production.
Introduction to Graph Machine Learning
Hugging Face provides a comprehensive overview of Graph Machine Learning, detailing how graphs are represented and the evolution from pre-neural features to Graph Neural Networks and Graph Transformers.
AI for Game Development: Creating a Farming Game in 5 Days (Part 1)
Hugging Face demonstrates how to use Stable Diffusion to establish a visual art style and concept art for a farming game, which is then implemented in Unity.
Accelerating PyTorch Transformers with Intel Sapphire Rapids – part 1
Hugging Face shows how to train PyTorch Transformers on a cluster of Intel Sapphire Rapids CPUs using IPEX and oneCCL, achieving up to 8× speed‑up over Ice Lake and near‑linear scaling across four nodes.
OpenAI Creating Next-Gen Characters Announcement
The source material provided for the announcement regarding creating next-gen characters is unavailable due to a 404 error, providing no technical details or official claims.
OFASys: A Framework for Multimodal Multitask Learning
Qwen introduces OFASys, an AI framework that simplifies multimodal multitask learning by allowing users to define complex tasks and modalities via a single-line Instruction interface.
Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese
Qwen has released Chinese CLIP, a vision-language model designed to overcome the cultural and linguistic limitations of English-centric CLIP models in cross-modal retrieval and image classification.
OpenAI Applied AI Research: Perspectives on AGI, Safety, and Continuous Learning
Lilian Weng of OpenAI's Applied AI Research team discusses the path toward AGI, the critical importance of model alignment and safety, and the role of continuous learning in AI development.
Zero-shot image segmentation with CLIPSeg
Hugging Face introduces CLIPSeg, a zero-shot image segmentation model that uses CLIP embeddings to create segmentation masks from either text or image prompts without requiring category-specific training.
Hugging Face Model Cards Documentation Framework
Hugging Face has released a suite of tools and resources, including a GUI-based creator tool and a standardized template, to improve the accessibility and standardization of machine learning model documentation.
Anthropic Research: Discovering Language Model Behaviors with Model-Written Evaluations
Anthropic researchers developed a method to automatically generate high-quality evaluations using language models to discover novel behaviors, including inverse scaling and sycophancy in larger models.
OpenAI Point-E: Generating 3D Point Clouds from Text Prompts
OpenAI has introduced Point-E, a system that generates 3D point clouds from complex text prompts in 1-2 minutes on a single GPU, prioritizing sampling speed over absolute sample quality.
OpenAI text-embedding-ada-002 release
OpenAI announced the text-embedding-ada-002 model, a unified, faster, cheaper embedding model that outperforms previous models on most tasks.
Hugging Face Audio Datasets Guide – How to Load, Process, and Stream Audio Data
Hugging Face announced a comprehensive guide showing that the 🤗 Datasets library can load, preprocess, and stream any audio dataset from the Hub with just a few lines of Python code, enabling efficient research on speech and audio tasks.
Hugging Face Ethics and Society Newsletter #2: Addressing Bias in Machine Learning
Hugging Face outlines a sociotechnical framework for mitigating machine learning bias by treating biases as risk factors that must be addressed across task definition, dataset curation, and model training.
Constitutional AI: Harmlessness from AI Feedback
Anthropic introduces Constitutional AI, a method for training harmless AI assistants using a set of rules (a constitution) and AI feedback instead of human labels to reduce harmful outputs.
Habana Gaudi2 vs Nvidia A100 80GB Performance Benchmarks
Hugging Face benchmarks show that Habana Gaudi2 provides approximately twice the throughput of Nvidia A100 80GB for both training and inference across BERT, Stable Diffusion, and T5-3B models.
Hugging Face Announces Bumblebee: Transformers and Stable Diffusion in Pure Elixir
Hugging Face released Bumblebee, a pure‑Elixir implementation of Transformers that brings models from GPT‑2 to Stable Diffusion to the Elixir ecosystem, enabling native CPU/GPU inference without external dependencies.
Illustrating Reinforcement Learning from Human Feedback (RLHF)
Hugging Face explains Reinforcement Learning from Human Feedback (RLHF), a three-step process used to align large language models with complex human values by optimizing them using a reward model based on human preferences.
OpenAI Supercomputing Infrastructure and Backend Systems Engineering
OpenAI engineer Jake Stangel describes the technical challenges of managing billion-dollar supercomputing clusters where extreme scale reveals hardware and kernel issues unseen by other users.
Deep Learning with Proteins – Hugging Face guide to protein language models and folding
Hugging Face announced a tutorial series showing how to fine‑tune protein language models and use ESMFold for protein folding, demonstrating that transfer learning techniques from NLP can be applied directly to protein sequence tasks.
Time Series Transformer probabilistic forecasting with 🤗 Transformers
Hugging Face released a vanilla Transformer model for global probabilistic time‑series forecasting, demonstrating state‑of‑the‑art performance on the Tourism Monthly benchmark.
Stable Diffusion Core ML on Apple Silicon – How to Run and Optimize
Hugging Face released Core ML‑converted Stable Diffusion checkpoints for Apple Silicon, enabling on‑device image generation in Python or Swift with up to 18 seconds per image on an M1 Max.
OpenAI Introducing ChatGPT
OpenAI has released ChatGPT, a conversational AI model based on the GPT-3.5 series, trained using Reinforcement Learning from Human Feedback (RLHF) to interact in a dialogue format.
VQ-Diffusion: Conditional Latent Diffusion in Discrete Space
VQ-Diffusion is a conditional latent diffusion model that operates on a quantized discrete latent space, offering faster inference and higher image quality than traditional autoregressive models.
Hugging Face 2023 Internship Program
Hugging Face has announced its 2023 internship program, offering roles across Open Source, Science, and Social Impact teams to democratize responsible machine learning.
Hugging Face Diffusion Models Class and Community Event
Hugging Face announced a free Diffusion Models Class launching November 28, 2022, accompanied by a live community event on November 30 featuring researchers from Stability AI, Meta, and Runway.