✷ The archive · 11 labs · 3,061 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.
GaLore: Advancing Large Model Training on Consumer-grade Hardware
GaLore enables the training of billion-parameter models on consumer-grade GPUs by reducing optimizer state memory requirements by over 82.5% through low-rank gradient projection.
Cosmopedia: Large-Scale Synthetic Data for LLM Pre-training
Hugging Face introduces Cosmopedia, the largest open synthetic dataset for LLM pre-training, containing 30 million files and 25 billion tokens generated by Mixtral-8x7B-Instruct-v0.1.
Phi-2 on Intel Meteor Lake: Local LLM Inference
Hugging Face demonstrates how to run the Microsoft Phi-2 model locally on Intel Meteor Lake (Core Ultra) processors using 4-bit quantization via OpenVINO and Optimum Intel.
Accenture, AWS, and Anthropic Collaboration for Enterprise AI Scaling
Anthropic, AWS, and Accenture have partnered to help organizations, particularly in regulated sectors, move generative AI from concept to production with a focus on security, reliability, and data privacy.
Claude 3 Models General Availability on Vertex AI
Anthropic has made Claude 3 Haiku and Claude 3 Sonnet generally available on Google Cloud's Vertex AI platform, allowing enterprises to scale generative AI solutions with enterprise-grade security.
Holiday Extras ChatGPT Enterprise Implementation
Holiday Extras deployed ChatGPT Enterprise across its organization, saving over 500 hours per week and achieving an estimated $500k in annual savings.
OpenAI and Salesforce Integration for Enterprise Trust and Safety
OpenAI provides the foundational generative AI models for the Salesforce Einstein AI Platform, enabling secure, enterprise-ready AI capabilities across Sales, Service, and Commerce Clouds.
Superhuman AI Integration with OpenAI API
Superhuman has integrated OpenAI's API to launch a suite of AI-powered email features that double the speed of inbox processing for its users.
Quanto: a PyTorch quantization backend for Optimum
Hugging Face introduces Quanto, a versatile and device-agnostic PyTorch quantization backend for Optimum designed to simplify low-precision model deployment across any modality.
Hugging Face Train on DGX Cloud
Hugging Face launched Train on DGX Cloud, a no-code service for Enterprise Hub organizations to fine-tune open models using NVIDIA H100 and L40S GPUs.
xAI Grok-1 Open Release
xAI has released the base model weights and network architecture of Grok-1, a 314 billion parameter Mixture-of-Experts model, under the Apache 2.0 license.
WebSight Dataset and Sightseer Model
Hugging Face introduced WebSight, a synthetic dataset of 2 million screenshot-to-HTML pairs, and Sightseer, a vision-language model capable of converting web screenshots into functional HTML code.
CPU Optimized Embeddings with Optimum Intel and fastRAG
Hugging Face and Intel have introduced a method to accelerate embedding models on Xeon CPUs using Optimum Intel and fastRAG, achieving up to 4.5x latency reduction and 4x throughput improvement via int8 quantization.
Ollama AMD Graphics Card Support Preview
Ollama has introduced preview support for AMD graphics cards on Windows and Linux, enabling hardware acceleration for all Ollama features across a wide range of Radeon, Radeon PRO, and Instinct series cards.
OpenAI Global News Partnerships with Le Monde and Prisa Media
OpenAI has partnered with Le Monde and Prisa Media to integrate their news content into ChatGPT, providing users with attributed summaries and direct links to original articles.
Healthify and OpenAI Collaboration for AI Health Coaching
Healthify has integrated OpenAI's GPT-4 Vision, GPT-4 Turbo, and Embeddings models to scale personalized nutrition and fitness coaching, resulting in a 50% increase in food tracking engagement.
Claude 3 Haiku Release Notes
Anthropic has released Claude 3 Haiku, a high-speed, affordable model designed for enterprise applications, featuring state-of-the-art vision capabilities and high token processing speeds.
OpenAI Governance Review and Leadership Confirmation
OpenAI has completed an independent review by WilmerHale, confirming full confidence in Sam Altman and Greg Brockman's leadership and implementing significant governance reforms.
OpenAI Board of Directors Expansion
OpenAI has appointed Dr. Sue Desmond-Hellmann, Nicole Seligman, and Fidji Simo to its Board of Directors, while CEO Sam Altman has rejoined the board.
Anthropic "Reflections on Qualitative Research" – Why Interpretability Needs a Qualitative Lens
Anthropic’s “Reflections on Qualitative Research” argues that interpretability work on AI models should prioritize qualitative methods and offers heuristics for judging such research.
Lifespan Healthcare System Uses GPT-4 to Improve Patient Health Literacy
Lifespan, Rhode Island's largest healthcare system, has deployed GPT-4 to simplify complex surgical consent forms from a college reading level to a middle school level to improve patient understanding and well-being.
Match Group Adopts ChatGPT Enterprise to Enhance Productivity and Product Development
Match Group has deployed ChatGPT Enterprise to thousands of employees to improve cross-functional collaboration, engineering efficiency, and product usability testing, while leveraging OpenAI's API for customer-facing features.
Paradigm Integrates GPT-4 to Accelerate Clinical Trial Patient Matching
Paradigm uses OpenAI's GPT-4 API to automate the evaluation of medical records for clinical trial eligibility, increasing accuracy by 10% and enabling the screening of hundreds of patients per minute.
OpenAI and Elon Musk: Clarification of Partnership and Mission
OpenAI released a statement clarifying its historical relationship with Elon Musk, the necessity of its for-profit transition to fund AGI development, and its commitment to broad benefit.
ConTextual: Benchmarking Multimodal Reasoning in Text-Rich Scenes
Hugging Face and UCLA researchers have introduced ConTextual, a dataset and leaderboard designed to evaluate how Large Multimodal Models (LMMs) jointly reason over text and visual cues in complex, text-rich images.
Hugging Face and Argilla Enable Collective Community Dataset Building
Hugging Face and Argilla have introduced a streamlined workflow using Hugging Face Spaces and Argilla to allow communities to collectively build high-quality, open-source datasets.
Claude 3 Model Family Release
Anthropic has released the Claude 3 model family, comprising Opus, Sonnet, and Haiku, which introduce advanced vision capabilities, improved accuracy, and new industry benchmarks in cognitive tasks.
Text-Generation Pipeline on Intel Gaudi 2 AI Accelerator
Hugging Face introduces a custom text-generation pipeline for Intel Gaudi 2 AI accelerators, enabling streamlined deployment of Llama 2 models (7b, 13b, and 70b) via Optimum Habana.
Anthropic Prompt Engineering for Business Performance
Anthropic outlines key prompt engineering techniques—including step-by-step reasoning, few-shot prompting, and prompt chaining—to improve Claude's accuracy, consistency, and cost-efficiency in business applications.
StarCoder2 and The Stack v2 Release
BigCode has released StarCoder2, a family of transparently trained open code LLMs in 3B, 7B, and 15B parameter sizes, powered by the massive new Stack v2 dataset.
Hugging Face TTS Arena: Benchmarking Text-to-Speech Models
Hugging Face has launched the TTS Arena, a crowdsourced, side-by-side comparison tool and leaderboard using an Elo rating system to objectively measure text-to-speech model quality.
Mistral AI Le Chat Release
Mistral AI has released Le Chat, a multilingual conversational assistant that serves as a demonstration of Mistral models' capabilities for both individual users and business environments.
Mistral Large and Mistral Small Release
Mistral AI has released Mistral Large, a flagship model with top-tier reasoning and multilingual capabilities, and Mistral Small, a model optimized for low-latency workloads.
AI Watermarking 101: Tools and Techniques
Hugging Face provides a comprehensive overview of AI watermarking techniques across images, text, and audio to combat deepfakes and ensure content provenance.
Introduction to Matryoshka Embedding Models
Matryoshka Embedding models allow for variable-size embeddings that can be truncated without significant performance loss, enabling a flexible trade-off between storage, speed, and accuracy.
Fine-Tuning Gemma Models in Hugging Face
Hugging Face provides a guide on using Parameter-Efficient Fine-Tuning (PEFT) and Low-Rank Adaptation (LoRA) to customize Google's Gemma models on GPUs and Cloud TPUs.
Hugging Face and Haize Labs Introduce Red-Teaming Resistance Leaderboard
Hugging Face and Haize Labs have launched the Red-Teaming Resistance (RTR) Benchmark to evaluate LLM robustness against high-quality, human-like adversarial prompts across specific safety violation categories.
Google Gemma Open LLM Release
Google has released Gemma, a family of open-access large language models based on Gemini, available in 2B and 7B parameter sizes with base and instruction-tuned variants.
Open Ko-LLM Leaderboard
Hugging Face and Upstage have launched the Open Ko-LLM Leaderboard to provide a fair, transparent evaluation ecosystem for Korean Large Language Models using private test sets to prevent contamination.
Hugging Face PEFT New LoRA Merging Methods
Hugging Face has introduced several new merging methods to the PEFT library, enabling users to combine multiple LoRA adapters from the same base model on the fly to synthesize new capabilities.
Synthetic Data with Open-Source LLMs Cuts Cost, Latency, and Carbon for Custom Models
Hugging Face shows how using open‑source LLMs to generate synthetic data and fine‑tune a small RoBERTa model reduces inference cost from $3061 to $2.7, latency from seconds to 0.13 s, and CO₂ emissions from ~1 t to 0.12 kg while matching GPT‑4 accuracy on financial sentiment classification.
Anthropic Election Integrity Strategy 2024
Anthropic has implemented a three-pronged strategy involving strict usage policies, adversarial red-teaming, and authoritative redirects to prevent the misuse of Claude during 2024 global elections.
OpenAI Sora: Video Generation Models as World Simulators
OpenAI has introduced Sora, a diffusion transformer model capable of generating high-fidelity video up to one minute long, suggesting that scaling video generation is a path toward general-purpose physical world simulators.
Ollama Windows Preview Release
Ollama has launched a preview version for Windows, enabling users to run, pull, and create large language models locally with native GPU acceleration and API access.
OpenAI Disrupts State-Affiliated Threat Actors Using AI
OpenAI, in partnership with Microsoft Threat Intelligence, terminated accounts of five state-affiliated threat actors from China, Iran, North Korea, and Russia who used AI for limited cybersecurity tasks.
AMD Pervasive AI Developer Contest
AMD and Hugging Face have partnered to launch the Pervasive AI Developer Contest, offering developers free access to AMD hardware and cash prizes to build AI applications in Generative AI, Robotics AI, and PC AI.
ChatGPT Memory and New User Controls
OpenAI has introduced a memory feature for ChatGPT that allows the model to remember details across conversations to provide more personalized responses without requiring repeated information.
Hugging Face TGI Messages API Release
Hugging Face has introduced a Messages API for Text Generation Inference (TGI) starting with version 1.4.0, enabling OpenAI Chat Completion API compatibility for open LLMs.
Ollama OpenAI Compatibility Update
Ollama has introduced built-in compatibility with the OpenAI Chat Completions API, allowing users to run local models using existing OpenAI-compatible tools and libraries.
Qwen1.5 Release Notes / What's New
Qwen has released Qwen1.5, a series of open-source base and chat models ranging from 0.5B to 110B parameters, featuring improved human alignment, multilingual capabilities, and native Hugging Face transformers integration.