A fashion photoshoot of a runway look inspired by blueprints. A large screen displays the word “Open Source” –ar 4:3 –style raw
Cohere
Nvidia and Salesforce double down on AI startup Cohere in $450 million round, source says | Reuters
“We’re excited to launch our startup program to empower early-stage AI innovation solving real world business challenges! Apply here:
“We’ve released a library of useful cookbooks to help you get started building powerful AI applications, like agents, with RAG and semantic search, with Cohere’s frontier models to tackle enterprise-grade workloads! Find them here:
Grok
“xAI’s Grok is rumored to be receiving two new modes, Socrates and D.E.I. This adds to the current selections of Normal and Fun to augment the AI chatbot’s responses
“NEWS: xAI might be working on ‘Socrates’ mode for Grok.
Hugging Face
“Thanks Jensen. Now you can use @nvidia NIM directly from the @huggingface hub for Llama3. Very excited to see that Hugging Face is becoming the gateway for AI compute!
“Optimum-NVIDIA from @nvidia – By changing just a single line of code, you can unlock up to 28x faster inference and 1,200 tokens/second on the NVIDIA platform. 🔥 📌 Optimum-NVIDIA is the first Hugging Face inference library to benefit from the new float8 format supported on
“Robotics + open source = <3 Hugging Face and Pollen Robotics show off first project: an open source robot that does chores
“MMUPD is now hosted on @huggingface Hub as a leaderboard 🏆 amazing to see LLaVA-1.6 is outperforming proprietary models on many subtasks 🤩 link in the next one!
“Summer blues? Cheer yourself (or a friend) up with the ComplimentBot 💖 I bootstrapped this app with the @Gradio javascript client and @huggingface ZeroGPU. If you’re building with AI, keep Gradio and Hugging Face Hub/spaces close by in your toolbox!
FineWeb: decanting the web for the finest text data at scale – a Hugging Face Space by HuggingFaceFW
“Hugging Face Embedding Container for @awscloud SageMaker is now generally available! Improve and simplify your embedding creation for Retrieval-Augmented Generation (RAG) applications. 🚀 What’s new 🆕: 🔥 Supports popular architectures: BERT, RoBERTa, XLM-RoBERTa, @nomic_ai
“Qwen2 released! 5 Sizes and multilingual in 29 languages. Check out on @huggingface! Blog:
Meta/Llama
“Thanks Jensen. Now you can use @nvidia NIM directly from the @huggingface hub for Llama3. Very excited to see that Hugging Face is becoming the gateway for AI compute!
“Last Week: Groq exceeded 30,000 Tokens / second input rate on Llama3 8B❗️ This Week: Llama3 70B at 40,792 Tokens/s input rate‼️ – FP16 Multiply, FP32 Accumulate – 7989 tokens in – full Llama context length Next Week: …? 😮
“A GPT-4 level chatbot, available to use completely free, running at over 800 tokens per second on Groq. I’m genuinely mindblown by LlaMA 3. Try it with the link in the next tweet.
“Put this mind bending achievement in perspective: @GroqInc runs Llama 70b in lossless precision on ~4 Wikipedia articles in quite literally the blink of an eye. – A 70B model in 16-bit precision with 32-bit accumulation (loss-less). – Processing ~8000 tokens in 0.2 seconds (or”
Mistral
“Run Mixtral with TensorRT-LLM with Parallelism Modes ✨ 📌 Mixture of Experts supports two parallelism modes, these are Expert Parallelism (EP) and Tensor Parallelism (TP). 📌 In TP mode (default) expert weight matrices are sliced evenly between all GPUs, so that all GPUs work
“Mistral fine-tuning API is out ! You can now fine-tune your own Mistral models and deploy them efficiently on La Plateforme :
“Breaking: First live demo of the @MistralAI fine-tuning API (released a few hours ago) is here: @sophiamyang walks through: – How to prep & validate your data – Hyper params – The fine-tuning API – Integrations (W&B, etc) – A treasure trove of collab notebooks & docs
My Tailor is Mistral | Mistral AI | Frontier AI in your hands
Phi
“Introducing Phi-3 WebGPU, a private and powerful AI chatbot that runs locally in your browser, powered by 🤗 Transformers.js and onnxruntime-web! 🔒 On-device inference: no data sent to a server ⚡️ WebGPU-accelerated (> 20 t/s) 📥 Model downloaded once and cached Try it out! 👇
“Phi-3 Medium (14B) and Small (7B) models are on the @lmsysorg leaderboard! 😍 Medium ranks near GPT-3.5-Turbo-0613, but behind Llama 3 8B. Phi-3 Small is close to Llama-2-70B, and Mistral fine-tunes. This proves that we cannot purely optimize for academic benchmarks. We need to
“Truly impressed by this: – real-time speech recognition – running locally in your browser – which means total confidentiality — 0 data sent to anyone – 100 different languages
Qwen
“Qwen2 released! 5 Sizes and multilingual in 29 languages. Check out on @huggingface! Blog:
“Qwen2 is the most impactful open LLM release since Llama 3! @Alibaba_Qwen just released their new multilingual model family, outperforming @AIatMeta Llama 3 🤯 Qwen2 comes in 5 sizes and is trained in 29 languages, achieving state-of-the-art performance across academic and chat
“💗Hello Qwen2! Happy to share the Qwen2 models to you all! 📖 BLOG:
Hello Qwen2 | Qwen
Generalizing an LLM from 8k to 1M Context using Qwen-Agent | Qwen
“After months of efforts, we are pleased to announce the evolution from Qwen1.5 to Qwen2. This time, we bring to you: ⭐ Base and Instruct models of 5 sizes, including Qwen2-0.5B, Qwen2-1.5B, Qwen2-7B, Qwen2-57B-A14B, and Qwen2-72B. Having been trained on data in 27 additional
“🔩Small but Strong: Qwen2-7B-Instruct
“We just got a new #1 open LLM!
Other Open Source News
“The Big AI Debate explained by Forbes. What is more beneficial or more dangerous: open source AI or proprietary controlled by 3 or 4 big players? The people who worry most about AI safety also tend to be he ones who overestimate the power of AI.”
The future of foundation models is closed-source
if the centralizing forces of data and compute hold, open and closed-source AI cannot both dominate long-term
“At COMPUTEX 2024, @nvidia CEO Jensen showing how Pandas code is now 50x faster on @GoogleColab after integration with RAPIDS cuDF over standard pandas. This is with zero code changes, and is available in the default runtime environment. Just add in %load-ext cudf.pandas over
“Open-source robotics is the way 🦾”
“The US is going to lose its leadership in AI if it doesn’t support more open research and open-source AI!”
“We have released GLM-4-520 and have the open-sourced version GLM-4-9B with superior performance beyond Llama-3-8B.
“Is Fineweb-edu the best open text dataset ever released? A big step in empowering all companies to train their own GPT5!
” 🤖Introducing 📺𝗢𝗽𝗲𝗻-𝗧𝗲𝗹𝗲𝗩𝗶𝘀𝗶𝗼𝗻: a web-based teleoperation software! 🌐Open source, cross-platform (VisionPro & Quest) with real-time stereo vision feedback. 🕹️Easy-to-use hand, wrist, head pose streaming. Code:
“The “weight” is nearly over! Today, at @ComputexTaipei, our Co-CEO, @chrlaf, officially announced the open release date of Stable Diffusion 3 Medium for June 12th. 🔗Sign up to the waitlist to be the first to know when the model releases:
Stable Audio Open — Stability AI
What I learned from looking at 900 most popular open source AI tools
An entirely open-source AI code assistant inside your editor · Ollama Blog
One clear, sad conclusion that we have to draw from our AIW study (https://arxiv.org/abs/2406.02061) is that all current SOTA open-weight LLMs that claim strong performance (eg Command-R+, Mistral, Dbrx-Instruct, Llama 3, Qwen, etc) are in fact seriously flawed in simple basic reasoning.
“New open benchmark released with 96% correlation to @lmsysorg Chatbot Arena with < $1 to run. 🤯 MixEval & MixEval-Hard combines existing benchmarks with real-world user queries from the web to close the gap between academic and real-world use! 👀 MixEval-Hard is a challenging
“The future of AI glasses is normal looking, light weight and affordable – meet Frame, AI Glasses by @brilliantlabsAR It is shipping to hackers and creators already. Frame is open-source platform with mic, camera, AR display. It leverages your phone (connectivity & audio) and
“there we go. please forgo your naive belief that Meta (or any company) will open source powerful AI. we are inevitably headed towards AGI monarchy.”
This week’s executive overview and top links are here:
AI News #36: Week Ending 06/07/2024 with Executive Summary and Top 40 Links
The post you just read is an deep dive extension of my weekly newsletter, This Week In AI, an executive summary of the top things to know in AI. Each week, I create an accessible overview for laypeople to feel confident they are conversant with the week’s AI developments. I include a curated list of must-click links of the week, to offer everyone a hands-on opportunity to explore the most intriguing updates in artificial intelligence across various categories, including robotics, imagery, video, AR/VR, science, ethics, and more. Beyond the overview, I post these topic-based deeper dives (below). If you haven’t read this week’s overview, I recommend starting there.
- Agents/Copilots
- Amazon
- Apple
- Artificial General Intelligence (AGI)
- Augmented and Virtual Reality (AR/VR)
- Autonomous Vehicles
- AI Audio
- Business and Enterprise AI
- Chips and Hardware
- Consumer Products
- Education
- Ethics/Legal Security
- Images/Photos
- International AI News
- Locally Run AI Models
- Mobile
- Meta
- Microsoft
- OpenAI
- Open Source
- Podcasts/YouTube
- Publishing and News
- Retrieval-Augmented Generation (RAG) News
- Robots and Embodiment
- Science and Medicine
- Video
- Vision/Multimodality
- X/Twitter/Grok
- Tech and Development
Credits/Sources
Most of these weekly links come from just a few prolific oversharing sources. Please follow them, as they work hard to find the news each week and they make it a lot easier for me to compile.
- Robert Scoble: https://x.com/Scobleizer
- Ethan Mollick: https://www.linkedin.com/in/emollick/
- Alan Thompson: https://lifearchitect.ai/
- Theoretically Media: https://www.youtube.com/@TheoreticallyMedia
- The Rundown: https://www.therundown.ai/
- Bilawal Sidhu: https://twitter.com/bilawalsidhu/
- TLDR: https://tldr.tech/ai
- Jeremiah Owyang: https://twitter.com/jowyang
- Nick St. Pierre: https://twitter.com/nickfloats
- Dr. Jim Fan: https://twitter.com/DrJimFan
- All About AI: https://www.youtube.com/@AllAboutAI
- Marshall Kirkpatrick: https://aitimetoimpact.com/
- AI News (Smol Talk): https://buttondown.email/ainews/archive/
For previous issues, please visit the archives!
Thanks for reading!
Leave a Reply