The archive · 11 labs · 3,062 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.

2701

Japanese Stable Diffusion release by rinna

rinna released Japanese Stable Diffusion, a Japanese‑language fine‑tuned version of Stable Diffusion that generates culturally appropriate images from Japanese prompts.

2702

Hugging Face Zero-Shot Evaluation on the Hub

Hugging Face has introduced zero-shot evaluation for causal language models on the Hub, enabling users to benchmark models up to 66 billion parameters without writing code.

2703

OpenAI DALL·E Beta Now Available Without Waitlist

OpenAI has removed the waitlist for the DALL·E beta, allowing immediate sign-up for a system that currently supports over 1.5 million users creating more than 2 million images daily.

2704

Hugging Face AutoTrain Image Classification

Hugging Face has added Image Classification to AutoTrain, enabling users to train custom image categorization models without writing code or configuring hyperparameters.

2705

Hugging Face Accelerate: Running Large Models with PyTorch

Hugging Face Accelerate enables the execution of massive AI models on consumer hardware by leveraging PyTorch's meta device and sharded checkpoints to manage memory across GPUs, CPU RAM, and disk.

2706

SetFit: Efficient Few-Shot Learning Without Prompts

Hugging Face introduces SetFit, a prompt-free framework for few-shot fine-tuning of Sentence Transformers that achieves high accuracy with minimal labeled data.

2707

Hugging Face Ethics and Society Newsletter #1

Hugging Face introduces its Ethics and Society newsletter and outlines a decentralized, value-driven approach to operationalizing AI ethics through collaboration, transparency, and responsibility.

2708

OpenAI Whisper Release

OpenAI has released Whisper, an automatic speech recognition (ASR) system trained on 680,000 hours of multilingual and multitask supervised data to improve robustness across accents, noise, and languages.

2709

Incredibly Fast BLOOM Inference with DeepSpeed and Accelerate

Hugging Face demonstrates sub‑millisecond per‑token generation for the 176B‑parameter BLOOM model using DeepSpeed‑Inference tensor parallelism and Accelerate pipeline parallelism on 8×80 GB A100 GPUs.

2710

Anthropic Toy Models of Superposition Research

Anthropic researchers use small ReLU networks to demonstrate that models can represent more features than they have dimensions through a phenomenon called superposition, provided those features are sparse.

2711

Diffusers 0.3 release adds image‑to‑image, textual inversion, inpainting, GPU optimizations, Mac MPS, ONNX support and new docs

Hugging Face announced Diffusers version 0.3, introducing image‑to‑image, textual inversion, experimental inpainting, smaller‑GPU optimizations, Mac MPS support, an ONNX exporter, expanded documentation, and a wave of community projects.

2712

Training Decision Transformers for Offline Reinforcement Learning

Hugging Face provides a guide and implementation for training an offline Decision Transformer from scratch to solve the HalfCheetah environment using the transformers Trainer and a custom data collator.

2713

Training Language Models with Megatron-LM

Hugging Face provides a guide on using NVIDIA's Megatron-LM framework to efficiently pre-train large language models on GPUs, including integration with the Transformers library.

2714

DALL·E Outpainting Feature Release

OpenAI has introduced Outpainting for DALL·E, allowing users to extend images beyond their original borders using natural language descriptions while maintaining visual consistency.

2715

OpenRAIL: Towards open and responsible AI licensing frameworks

Hugging Face announced OpenRAIL, a set of AI‑specific licenses that combine open access with use‑based restrictions to promote responsible deployment of machine‑learning models.

2716

OpenAI Announces Alignment Strategy Using RL from Human Feedback and InstructGPT

OpenAI detailed its alignment approach, highlighting RL from human feedback and InstructGPT as effective, low‑cost methods for making language models follow human intent while acknowledging current limitations and future goals for AI‑driven alignment research.

2717

Visualize proteins on Hugging Face Spaces

Hugging Face provides a guide on integrating 3Dmol.js into Hugging Face Spaces via Gradio to enable 3D protein structure visualization in the browser.

2718

Pre-training BERT with Hugging Face Transformers and Habana Gaudi

Hugging Face demonstrates how to pre-train BERT-base from scratch using Habana Gaudi DL1 instances on AWS, achieving a 25% cost reduction compared to NVIDIA V100 GPU-based training.

2719

Stable Diffusion with Diffusers

Hugging Face announced how to run the text-to-image Stable Diffusion model using the Diffusers library, providing code, licensing details, and explanations of the model’s latent diffusion architecture.

2720

Anthropic Red Teaming Language Models Report – Methods, Scaling Behaviors, and Lessons Learned

Anthropic released a detailed study showing that reinforcement‑learning‑from‑human‑feedback (RLHF) models become harder to red‑team as they scale, while other model types show flat red‑teamability, and they published a 38,961‑attack dataset to help the community improve safety.

2721

Deploying Hugging Face ViT on Vertex AI

Hugging Face demonstrates how to deploy a Vision Transformer (ViT) model on Google Cloud's Vertex AI platform to achieve Kubernetes-level scalability with significantly less code.

2722

Vision Transformers on Hugging Face Optimum Graphcore

Hugging Face and Graphcore enable the efficient fine-tuning of Vision Transformer (ViT) models on Intelligence Processing Units (IPUs) using the Optimum library, demonstrated through a multi-label chest X-ray classification task.

2723

Hugging Face integrates LLM.int8() 8-bit matrix multiplication into Transformers and Accelerate

Hugging Face released an integration of the LLM.int8() 8-bit quantization technique into the Transformers and Accelerate libraries, enabling inference of very large language models with no performance degradation and up to 2× memory savings.

2724

Hugging Face's TensorFlow Philosophy

Hugging Face outlines its technical approach to TensorFlow integration, emphasizing deep alignment with Keras and XLA to optimize performance and deployment for transformer models.

2725

Skops library enables scikit-learn model hosting, documentation, and collaboration on Hugging Face Hub

Hugging Face announced Skops, a library that lets users package, document, and push scikit-learn models to the Hub with automatic model cards and inference widgets.

2726

Deploying Hugging Face ViT on Kubernetes with TF Serving

Hugging Face provides a technical guide on scaling Vision Transformer (ViT) deployments using Docker and Kubernetes with TensorFlow Serving to enable production-grade autoscaling and resource management.

2727

OpenAI Moderation Endpoint Release

OpenAI has released a free Moderation endpoint providing GPT-based classifiers to help developers detect and block prohibited content in their applications.

2728

Hugging Face Sentence Transformers Training Guide (historical reference)

Hugging Face announced an outdated tutorial for training and fine-tuning Sentence Transformers models, explaining model architecture, dataset formats, loss functions, and how to upload models, while directing users to newer training APIs.

2729

Hugging Face Deep RL Class Unit 8: Proximal Policy Optimization (PPO) Explained

Hugging Face introduced a detailed walkthrough of Proximal Policy Optimization (PPO), showing how clipping the policy‑update ratio stabilizes training and providing a from‑scratch PyTorch implementation tested on CartPole‑v1 and LunarLander‑v2.

2730

Hugging Face Private Hub Introduction

Hugging Face introduced the Private Hub (now Enterprise Hub), a secure, compliant environment for companies to collaborate on machine learning models, datasets, and spaces using a unified toolset.

2731

Nyströmformer: Approximating self-attention in linear time and memory via the Nyström method

Nyströmformer reduces the time and memory complexity of standard self-attention from O(n^2) to O(n) by using the Nyström method to approximate the softmax matrix via sampled landmarks.

2732

Hugging Face Recommendations for the U.S. National AI Research Resource

Hugging Face submitted a response to the White House and National Science Foundation advocating for the National AI Research Resource (NAIRR) to prioritize ethical expertise, standardized documentation, and accessible tooling to democratize AI research.

2733

OpenAI Announces Efficient Training of Language Models for Fill-in-the-Middle

OpenAI released a new training method that enables autoregressive language models to perform fill‑in‑the‑middle tasks without sacrificing left‑to‑right generation quality, offering a simple and efficient way to add infilling capability.

2734

Hugging Face Datasets Audio and Vision Documentation Update

Hugging Face has expanded the 🤗 Datasets library documentation to include dedicated guides and tools for audio and vision datasets, simplifying the process of loading and processing multi-modal data.

2735

Faster Text Generation with TensorFlow and XLA

Hugging Face has enabled XLA compilation for text generation in the transformers library using TensorFlow, resulting in speedups of up to 100x and performance that often exceeds PyTorch.

2736

A Hazard Analysis Framework for Code Synthesis Large Language Models

OpenAI has introduced a hazard analysis framework to identify and mitigate technical, social, political, and economic risks associated with the deployment of advanced code generation models like Codex.

2737

Deploying TensorFlow Vision Models in Hugging Face with TF Serving

Hugging Face provides a guide on deploying TensorFlow vision models, such as the Vision Transformer (ViT), using TensorFlow Serving to expose them as REST or gRPC endpoints.

2738

Advantage Actor Critic (A2C) Explained

Hugging Face explains Advantage Actor Critic (A2C), a hybrid reinforcement learning architecture that reduces variance in policy gradient estimation by combining policy-based and value-based methods.

2739

OpenAI DALL·E beta release

OpenAI announced that DALL·E is now available in beta for 1 million invited users, offering free monthly credits and commercial usage rights.

2740

Reducing Bias and Improving Safety in DALL·E 2

OpenAI has implemented a system-level technique for DALL·E 2 to increase the diversity of generated people and enhanced safety filters to prevent the creation of deceptive content.

2741

Dynamic Adversarial Data Collection (DADC) for MNIST – Hugging Face tutorial

Hugging Face announced a step‑by‑step tutorial for Dynamic Adversarial Data Collection (DADC) on MNIST, showing how to gather human‑generated adversarial examples and retrain models to improve robustness.

2742

DALL·E 2: Extending Creativity

OpenAI's DALL·E 2 research preview has been adopted by over 3,000 artists across 118 countries to integrate AI-generated imagery into diverse creative workflows ranging from medical support to high-end culinary arts.

2743

BLOOM 176B Training Technology Overview

Hugging Face released a detailed breakdown of how the 176‑billion‑parameter BLOOM model was trained using 384 A100 GPUs, Megatron‑DeepSpeed 3D parallelism, and BF16 mixed‑precision, highlighting the hardware, software stack, and engineering challenges.

2744

Building a Playlist Generator with Sentence Transformers

Hugging Face demonstrates how to create a semantic search-based playlist generator using Sentence Transformers for lyric embeddings and Gradio Blocks for the user interface.

2745

BLOOM: The World's Largest Open Multilingual Language Model

Hugging Face and the BigScience project have released BLOOM, a 176-billion parameter open-access multilingual language model trained in complete transparency across 46 natural languages and 13 programming languages.

2746

Anthropic Research: Language Models (Mostly) Know What They Know

Anthropic research demonstrates that large language models can effectively evaluate the validity of their own claims and predict their likelihood of knowing an answer, providing a path toward more honest AI.

2747

Getting Started with Sentiment Analysis on Twitter

Hugging Face provides a guide on using the Inference API to perform automated sentiment analysis on Twitter data for both developers and non-coders.

2748

Policy Gradient with PyTorch – Hugging Face Deep RL Class Unit 5

Hugging Face released a tutorial that walks through implementing the REINFORCE Monte‑Carlo policy‑gradient algorithm from scratch in PyTorch and evaluates it on CartPole‑v1, PixelCopter, and Pong.

2749

Getting Started with Machine Learning using Sentence Transformers

Hugging Face provides a guide for beginners to start their first ML project using the Sentence Transformers library to create embeddings and perform semantic search.

2750

DALL·E 2 Pre-training Mitigations

OpenAI implemented active learning for data filtering, a reweighting scheme to prevent bias amplification, and a clustering-based deduplication process to eliminate image regurgitation in DALL·E 2.