Hugging Face's CEO on Open Source AI, Model Routing, and the Future of Competition
As governments weigh new restrictions on frontier AI models, one question is becoming increasingly important: what role should open source play in the future of artificial intelligence?
Theo Jaffee and Sofia Puccini speak with Hugging Face CEO Clément Delangue about AI regulation, open source safety, model routing, and why he believes competition—not consolidation—is essential for the industry's future.
They discuss GPT-5, government oversight of frontier models, Hugging Face surpassing $100 million in annual recurring revenue, local AI, China's open-source ecosystem, Europe's AI ambitions, and why routing workloads across specialized models could fundamentally reshape where value is created in AI.
Resources:
Follow Clément Delangue on X: https://x.com/ClementDelangue
Follow Theo Jaffee on X: https://x.com/theojaffee
Follow Sofia Puccini on X: https://x.com/schisofrenia
Follow MTS on X: https://x.com/mtslive
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
Open source AI matters more than ever, according to Hugging Face's Clem Delangue
Open source AI is booming, according to Hugging Face CEO Clem Delangue. The company has grown into something like a GitHub for AI in recent years, where AI builders can share and download open models and datasets, now used by roughly half the Fortune 500. Delangue has seen the same story play out again and again: companies start out on frontier APIs, but as they scale, the costs push them towards open source models.
On this episode of TechCrunch's Equity podcast, Rebecca Bellan talked to Delangue about why the open vs closed source fight matters in the wake of Anthropic’s halted Fable release, and why he's worried about the possibility that a handful of big companies could end up controlling everything.
Listen to the full episode to hear more about:
How Chinese labs are producing the majority of open models being downloaded in the U.S., and why Delangue thinks that's a problem worth fixing rather than a reason to distrust open source itself.
How Hugging Face is choosing capital efficiency over the usual Silicon Valley fundraising playbook, including why the company turned down a large investment from Nvidia last year.
Why he sees robotics as an even more urgent case for open, transparent AI than chatbots or coding tools, given how much of your home and family life a robot ends up seeing.
Subscribe to Equity on YouTube, Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod.
Chapters:
00:00 Intro
00:33 Breaking down open source growth data
04:34 What's driving the open source resurgence
08:47 Who’s using Hugging Face, and how?
10:28 China overtakes the US in open model downloads
16:34 Safety, access, and the risk of AI power concentration
24:03 Hugging Face's approach to legal risk
28:00 Turning down Nvidia
31:47 Underinvested opportunities: local AI, bio, robotics
Learn more about your ad choices. Visit megaphone.fm/adchoices
Hugging Face's Clem Delangue on Open Source AI and the LLM Bubble | MTS Live
Clem Delangue joins MTS to discuss the global open-source AI landscape, the current large language model bubble, and the future of consumer robotics.
Originally aired on MTS, Theo Jaffee and Sofia Puccini speak with Clément Delangue, CEO at Hugging Face, about the global open-source AI race, why he believes the real bubble is in API-based large language models, and how robotics could become the next major interface for AI. They also discuss AI safety, U.S.-China competition, open-weight models, and why Hugging Face became the infrastructure layer for open AI development.
Resources:
Follow Clem on X: @ClementDelangue
Follow Theo on X: @theojaffee
Follow Sofia on X: @schisofrenia
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
We’re doing AI all wrong. Here’s how to get it right | Sasha Luccioni
Artificial intelligence is changing everything — but at what cost? AI sustainability expert Sasha Luccioni exposes how tech companies' massive data centers are burning through energy and wrecking the planet. She introduces a powerful alternative: small but mighty AI models that could flip the script and make the technology smarter, fairer and sustainable.
Hosted on Acast. See acast.com/privacy for more information.
(BNS) Hugging Face Founder Clément Delangue
Clem discusses his journey from early computing experiences to founding Hugging Face, emphasizing the importance of community, collaboration, and open-source technology in the AI landscape. He reflects on the evolution of technology, the significance of user feedback, and the need for a diverse range of AI models. Clem also shares insights on the startup ecosystem in Europe and the unique advantages of New York City for AI entrepreneurs.
Learn more about your ad choices. Visit megaphone.fm/adchoices
Challenging the Average With Open-Source AI: Hugging Face’s Thomas Wolf
Thomas Wolf is the cofounder and chief science officer of open-source AI platform Hugging Face, which provides access to thousands of pretrained AI models that can be downloaded and run locally. With over 10 million users, getting started on the site can be a daunting task. Thomas explains how the company aims to improve its accessibility through documentation on the company blog as well as community feedback, similar to social media likes and upvoting.
Thomas and Sam discuss the benefits and trade-offs of both open-source and closed-source AI models, as well as the evolution of microchips and the future of hardware and software development — as well as the hopes Thomas has for the future of coding with AI, starting with his children’s generation. Read the episode transcript here.
Guest bio:
Thomas Wolf is cofounder and chief science officer of Hugging Face, a collaborative AI platform. Wolf likes creating open-source software (OSS) that makes complex research, models, and data sets widely accessible. He can also be found pushing for open science in research in AI and machine learning, to try lowering the gap between academia and industrial labs through projects like the BigScience Workshop. He also writes and produces education content on AI, machine language, and natural language processing, including the reference book Natural Language Processing with Transformers, The Ultra-Scale Playbook, his blog, and videos.
Me, Myself, and AI is a podcast produced by MIT Sloan Management Review and hosted by Sam Ransbotham. It is engineered by David Lishansky and produced by Allison Ryder.
We encourage you to rate and review our show. Your comments may be used in Me, Myself, and AI materials.
Building the "App Store" for Robots: Hugging Face's Thomas Wolf on Physical AI
Thomas Wolf, co-founder and Chief Science Officer of Hugging Face, explains how his company is applying the same community-driven approach that made transformers accessible to everyone to the emerging field of robotics. Thomas discusses LeRobot, Hugging Face's ambitious project to democratize robotics through open-source tools, datasets, and affordable hardware. He shares his vision for turning millions of software developers into roboticists, the challenges of data scarcity in robotics versus language models, and why he believes we're at the same inflection point for physical AI that we were for LLMs just a few years ago.
Hosted by: Sonya Huang and Pat Grady, Sequoia Capital
Hugging Face’s co-founder on bringing open-source AI to life with cute robots
Hugging Face’s new AI robot, the Reachy Mini, has already racked up $1 million in sales just five days after launch. But the company isn’t trying to build a chore-doing humanoid just yet. Instead, Hugging Face sees the Reachy Mini as a hackable, desk-friendly device that's part entertainment, part entry point for developers and consumers to experiment with AI in physical form.
On this episode of Equity, co-founder Thomas Wolf joins to explain why open-source AI needs hardware, how Hugging Face is thinking about robotics long term, and what might happen if people actually start coding apps for their robots.
We'll also get into:
How Hugging Face plans to leap from software to hardware.
Hugging Face's ambitions to one day sell a full-sized humanoid robot.
The role of privacy in consumer robotics, and how open-source can address it.
Equity is TechCrunch’s flagship podcast, produced by Theresa Loconsolo, and posts every Wednesday and Friday.
Subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod.
Credits: Equity is produced by Theresa Loconsolo with editing by Kell. We’d also like to thank TechCrunch’s audience development team. Thank you so much for listening, and we'll talk to you next time.
Learn more about your ad choices. Visit megaphone.fm/adchoices
“The future is agents”: Building a platform for RAG agents
Contextual AI offers a platform for building RAG agents. Get started with their docs here.
Connect with Douwe on LinkedIn.
Congrats to Stack Overflow user Smrutiranjan Sahu, who earned a Stellar Question badge by asking How to define type for a function callback (as any function type, not universal any) used in a method parameter.
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.
Goodbye TikTok, Ni Hao RedNote? + A.I.'s Environmental Impact + Meta's Masculine Energy
The deadline for TikTok to sell or to face a ban is fast approaching. We discuss how Supreme Court justices — who opted on Friday to uphold the law — reacted to arguments in the case, whether the Chinese government might allow Elon Musk to buy the app, and why self-proclaimed TikTok refugees are rushing to a different Chinese app, called RedNote. Then, we talk with an A.I. industry insider about what we actually know about how bad artificial intelligence is for the environment. And finally, after Mark Zuckerberg’s recent appearance on Joe Rogan’s podcast, Casey offers Kevin some ideas for how to bring more “masculine energy” to Meta.
Guest:
Sasha Luccioni, A.I and climate lead at Hugging Face.
Additional Reading:
Supreme Court Backs Law Requiring TikTok to Be Sold or Banned
‘Red Note,’ a Chinese App, Is Dominating Downloads, Thanks to TikTok Users
China Weighs Sale of TikTok US to Musk as a Possible Option
Matter of Opinion Podcast: How Democrats Drove Silicon Valley Into Trump’s Arms
We want to hear from you. Email us at hardfork@nytimes.com. Find “Hard Fork” on YouTube and TikTok.
Subscribe today at nytimes.com/podcasts or on Apple Podcasts and Spotify. You can also subscribe via your favorite podcast app here https://www.nytimes.com/activate-access/audio?source=podcatcher. For more podcasts and narrated articles, download The New York Times app at nytimes.com/app.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
What's our relationship to AI? It's complicated | AC Coppens, Kasley Killam and Apolinário Passos
In a lively conversation from TED's brand-new Next Stage, social scientist Kasley Killam, technologist Apolinário Passos and futurist AC Coppens explore the intricate dynamics of human-AI relationships — and show how AI is already changing the ways we live, work and connect with each other.
Hosted on Acast. See acast.com/privacy for more information.
#218 Jeff Boudier: The Future of Open Source AI Development (Hugging Face)
This episode is sponsored by Oracle.
Oracle Cloud Infrastructure, or OCI is a blazing fast and secure platform for your infrastructure, database, application development, plus all your AI and machine learning workloads. OCI costs 50% less for compute and 80% less for networking. So you're saving a pile of money. Thousands of businesses have already upgraded to OCI, including MGM Resorts, Specialized Bikes, and Fireworks AI.
Cut your current cloud bill in HALF if you move to OCI now: https://oracle.com/eyeonai
In this episode of the Eye on AI podcast, Jeff Boudier, Head of Product and Growth at Hugging Face, joins Craig Smith to uncover how the platform is empowering AI builders and driving the open-source AI revolution.
With a mission to democratize AI, Jeff walks us through Hugging Face's journey from a chatbot for teens to the leading platform hosting over 1 million public AI models, datasets, and applications. We explore how Hugging Face is bridging the gap between enterprises and open-source innovation, enabling developers to build cutting-edge AI solutions with transparency and collaboration.
Jeff dives deep into Hugging Face's tools and features, from hosting private and public models to fostering a thriving ecosystem of AI builders. He shares insights on the transformative impact of technologies like Transformers, transfer learning, and no-code solutions that make AI accessible to more creators than ever before.
We also discuss Hugging Face's latest innovation, 'Hugs,' designed to help enterprises seamlessly integrate open-source AI within their infrastructure while retaining full control over their data and models.
Tune in to discover how Hugging Face is shaping the future of AI development, why open-source models are catching up with proprietary ones, and what trends are driving innovation across AI disciplines.
Don't forget to like, subscribe, and hit the notification bell for more!
Stay Updated:
Craig Smith Twitter: https://twitter.com/craigss
Eye on A.I. Twitter: https://twitter.com/EyeOn_AI
(00:00) Introduction to Jeff Boudier
(02:16) How Hugging Face Empowers AI Builders
(05:26) Transition from Chatbot to Leading AI Platform
(07:07) Hosting AI Models: Public and Private Options
(10:13) What Does Hosting Models on Hugging Face Mean?
(14:22) Hugging Face vs. GitHub: Key Differences
(19:09) Navigating 1 Million Models on the Hugging Face Hub
(22:33) Leaderboards and Filtering AI Models
(25:26) Building Applications with Hugging Face Models
(28:03) AI Innovation: From Code to Model-Driven Development
(30:45) Frameworks for Agentic Systems and Hugging Chat
(35:20) Open Source vs. Proprietary AI: The Future
(40:41) Introducing 'Hugs': Open AI for Enterprises
(44:59) The Role of No-Code in AI Development
(47:26) Hugging Face's Vision
Building the Open Source AI Revolution (with Hugging Face CEO, Clem Delangue)
We sit down with Hugging Face CEO Clem Delangue to understand the current state of the open source AI ecosystem. Hugging Face is the leading platform to host and collaborate on AI models, datasets, and applications. They also have a compute offering for AI builders to train their models directly on the platform. Clem has a contrarian take on the future: there will not be just a few major foundation model companies with everyone using their APIs. But rather, that thousands of companies will have their own specialized AI models built in-house for their particular use case. It's obviously a very dynamic landscape and we'll have to see how it shakes out, but Clem has a pretty great viewpoint to see it all, working with their 5 million registered Hugging Face users!
Links:
Hugging Face
Hugging Face's Series D valuing the company at $4.5 billion
Sponsors:
Sierra: https://bit.ly/acquiredsierra
Towards high-quality (maybe synthetic) datasets
As Argilla puts it: “Data quality is what makes or breaks AI.” However, what exactly does this mean and how can AI team probably collaborate with domain experts towards improved data quality? David Berenstein & Ben Burtenshaw, who are building Argilla & Distilabel at Hugging Face, join us to dig into these topics along with synthetic data generation & AI-generated labeling / feedback.
Sponsors:
Fly.io – The home of Changelog.com — Deploy your apps close to your users — global Anycast load-balancing, zero-configuration private networking, hardware isolation, and instant WireGuard VPN connections. Push-button deployments that scale to thousands of instances. Check out the speedrun to get started in minutes.
WorkOS – A platform that gives developers a set of building blocks for quickly adding enterprise-ready features to their application. Add Single Sign-On (Okta, Azure, Google, Microsoft OAuth), sync users from any SCIM directory, HRIS integration, audit trails (SIEM), free magic link sign-in. WorkOS is designed for developers and offers a single, elegant interface that abstracts dozens of enterprise integrations. Learn more and get started at WorkOS.com
Eight Sleep – Take your sleep and recovery to the next level. Go to eightsleep.com/PRACTICALAI and use the code PRACTICALAI to get $350 off your very own Pod 4 Ultra. You can try it for free for 30 days - but we’re confident you will not want to return it. Once you experience AI-optimized sleep, you’ll wonder how you ever slept without it. Currently shipping to: United States, Canada, United Kingdom, Europe, and Australia.
Featuring:
Ben Burtenshaw – GitHub, LinkedIn, X
David Berenstein – GitHub, LinkedIn, X
Chris Benson – Website, GitHub, LinkedIn, X
Daniel Whitenack – Website, GitHub, X
Show Notes:
Argilla
Distilabel
Synthetic Data Generator UI
Hugging Face and Argilla meetups
Upcoming Events:
Register for upcoming webinars here!
Energy Star Ratings for AI Models with Sasha Luccioni - #687
Today, we're joined by Sasha Luccioni, AI and Climate lead at Hugging Face, to discuss the environmental impact of AI models. We dig into her recent research into the relative energy consumption of general purpose pre-trained models vs. task-specific, non-generative models for common AI tasks. We discuss the implications of the significant difference in efficiency and power consumption between the two types of models. Finally, we explore the complexities of energy efficiency and performance benchmarking, and talk through Sasha’s recent initiative, Energy Star Ratings for AI Models, a rating system designed to help AI users select and deploy models based on their energy efficiency.
The complete show notes for this episode can be found at http://twimlai.com/go/687.
740: Local AI Models in JavaScript - Machine Learning Deep Dive With Xenova
Scott and Wes are joined by special guest Xenova to explore local AI models in JavaScript. From Hugging Face to Transformers.js and practical applications like real-time speech recognition and object detection, this episode dives deep into the world of machine learning.
Show Notes 00:00 Welcome to Syntax!
00:41 Brought to you by Sentry.io
01:05 Who is Xenova?
02:08 What is Hugging Face?
03:29 What is Transformers.js?
06:16 How was the library developed?
SponsorBlock
09:04 How is it able to run?
10:09 Do they have to run in Python and how does Onnx work?
Onnx.ai
Hugging Face Optimum
14:19 What are some things you can do with this tech?
16:15 Vision tools.
17:38 This is actually running locally.
18:35 Doodle Dash
21:09 They currently run on CPU, what is required to make it run on GPU?
24:44 Can you run in JavaScript?
28:32 How it works with image vectors.
34:23 Why would people want to run it in another language?
35:55 Resizing images in the browser instead of on the server.
38:55 Applications distributed on the web vs running locally.
43:54 Electron has Node and Chrome, where would you run Transformers.js?
44:32 The API of Transformers.js
46:30 Object Detection.
Semantic Image Search Client
Real-Time Object Detection
Background Removal Tool
48:33 What is the easiest way to get started?
51:26 Real-time speech recognition on the horizon?
52:08 Will we ever be able to run Stable Diffusion via JavaScript?
56:10 The Web LLM.
57:22 Practical applications for YouTube.
59:39 What we want to build for Syntax.fm.
01:06:43 Mean pooling, why it’s necessary.
01:09:30 Stopping YouTube spam comments.
01:10:34 K-Means Clustering.
Text Clustering
01:13:49 Quantization.
01:17:35 Sick Picks + Shameless Plugs.
Sick Picks Xeonva: WebGPU
Shameless Plugs Xenova: Xenova on X
Hit us up on Socials! Syntax: X Instagram Tiktok LinkedIn Threads
Wes: X Instagram Tiktok LinkedIn Threads
Scott:X Instagram Tiktok LinkedIn Threads
Randy: X Instagram YouTube Threads
How to train your own Large Multimodal Model — with Hugo Laurençon & Leo Tronchon of HuggingFace M4
Latent Space is heating up! Our paper club ran into >99 person Discord limits, oops.
We are also introducing 2 new online meetups: LLM Paper Club Asia for Asia timezone (led by Ivan), and AI in Action: hands-on application of AI (led by KBall).
To be notified of all upcoming Latent Space events, subscribe to our new Luma calendar (sign up for individual events, or hit the RSS icon to sync all events to calendar).
In the halcyon open research days of 2022 BC (Before-ChatGPT), DeepMind was the first to create a SOTA multimodal model by taking a pre-existing LLM (Chinchilla 80B - now dead?) and pre-existing vision encoder (CLIP) and training a “glue” adapter layer, inspiring a generation of stunningly cheap and effective multimodal models including LLaVA (one of the Best Papers of NeurIPS 2023), BakLLaVA and FireLLaVA.
However (for reasons we discuss in today’s conversation), DeepMind’s Flamingo model was never open sourced. Based on the excellent paper, LAION stepped up to create OpenFlamingo, but it never scaled beyond 9B. Simultaneously, the M4 (audio + video + image + text multimodality) research team at HuggingFace announced an independent effort to reproduce Flamingo up to the full 80B scale:
The effort started in March, and was released in August 2023.
We happened to visit Paris last year, and visited HuggingFace HQ to learn all about HuggingFace’s research efforts, and cover all the ground knowledge LLM people need to become (what Chip Huyen has termed) “LMM” people. In other words:
What is IDEFICS?
IDEFICS is an Open Access Visual Language Model, available in 9B and 80B model sizes. As an attempt to re-create an open-access version of Flamingo, it seems to track very well on a range of multimodal benchmarks (which we discuss in the pod):
You can see the reasoning abilities of the models to take a combination of interleaved images + text in a way that allows users to either describe images, ask questions about the images, or extend/combine the images into different artworks (e.g. poetry).
📷 From IDEFICS’s model card and blog post
The above demo screenshots are actually fine-tuned instruct versions of IDEFICS — which are again in 9B and 80B versions.
IDEFICS was built by connecting two unimodal models together to provide the multi-modality you see showcased above.
* Llama v1 for language (specifically huggyllama/llama-65b) - the best available open model at the time, to be swapped for Mistral in the next version of IDEFICS
* A CLIP model for vision (specifically laion/CLIP-ViT-H-14-laion2B-s32B-b79K - after a brief exploration of EVA-CLIP, which we discuss on the pod)
OBELICS: a new type of Multimodal Dataset
IDEFICS’ training data used the usual suspect datasets, but to get to par with Flamingo they needed to create a new data set.
Enter OBELICS: “An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents”:
* 115B text tokens
* 141M English documents
* 353M images
These bullets are carefully curated and filtered by going through Common Crawl dumps between FEB 2020 - FEB 2023. We discuss the 2 months of mindnumbing, unglamorous work creating this pipeline:
There’s a lot of mentions of ‘multi-modal' web documents’ which deserves some explanation. We’ll show you instead of tell you:
You can see from this graph that OBELICS ends up outperforming the other image-text pairs dataset (LAION in this case) when stacked head-to-head.
You can view a subset of OBELICS and perform visualizations on them here:
2024 Update: WebSight et al
Most of this interview was recorded on Halloween 2023 at HuggingFace’s headquarters in Paris:
In anticipation of an IDEFICS v2 release. However, several roadblocks emerged, including a notable scandal around CSAM in LAION-5B, which affected all models using that dataset. The M4 team have adopted a strategy of shipping smaller advancements in 2024, and the first ship of the year is WebSight, a dataset of 823,000 HTML/CSS codes representing synthetically generated English websites, each accompanied by a corresponding screenshot (rendered with Playwright). This is intended for tasks like screenshot-to-code workflows like Vercel’s V0 or TLDraw, and will be part of the dataset for IDEFICS-2.
As noted in our Best Papers recap, synthetic data is emerging as one of the top themes of 2024, and the IDEFICS/OBELICS team have wasted no time enabling themselves with it.
Timestamps
* [0:00:00] Intro
* [0:00:00] Hugo, Leo’s path into multimodality
* [0:09:16] From CLIP to Flamingo
* [0:12:54] Benchmarks and Evals
* [0:16:54] OBELICS dataset
* [0:34:47] Together Redpajama v2
* [0:37:12] GPT4 Vision
* [0:38:44] IDEFICS model
* [0:40:57] Query-Key Layernorm for training
* [0:46:40] Choosing smaller vision encoders - EVA-CLIP vs SIG-GLIP
* [0:49:02] IDEFICS v2
* [0:52:39] Multimodal Hallucination
* [0:59:12] Why Open Source Multimodality
* [1:05:29] Naming: M4, OBELICS, IDEFICS
* [1:08:56] 2024 Update from Leo
Show Notes
* Introducing IDEFICS: An Open Reproduction of State-of-the-Art Visual Language Model
* IDEFICS Knowledge sharing memo: technical lessons and mistakes
* Victor Sanh memo
* OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents
* Papers cited:
* BLOOM: A 176B-Parameter Open-Access Multilingual Language Model
* Barlow Twins: Self-Supervised Learning via Redundancy Reduction
* CLIP paper: Learning Transferable Visual Models From Natural Language Supervision
* Vision Transformers paper: An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
* Flamingo paper: a Visual Language Model for Few-Shot Learning
* April 2022 preprint from DeepMind, blogpost
* VQAV2 paper: Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering
* OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge (https://okvqa.allenai.org/)
* MMBench: Is Your Multi-modal Model an All-around Player?
* Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
* Sig-GLIP paper: Sigmoid Loss for Language Image Pre-Training
* Nougat: Neural Optical Understanding for Academic Documents
* MMC4 (Multimodal C4): An Open, Billion-scale Corpus of Images Interleaved With Text
* Dall-E 3 paper: Improving Image Generation with Better Captions
* GPT-4V(ision) system card from OpenAI
* Query-Key Layernorm trick: paper (Scaling Vision Transformers to 22 Billion Parameters), tweet
* EVA-CLIP: Improved Training Techniques for CLIP at Scale
* “We intially explored using a significantly bigger vision encoder (the biggest in open-access at that time) with EVA-CLIP. However, we ran into training instabilities very quickly. To lower the risks associated to the change of vision encoder, we decided to continue with laion/CLIP-ViT-H-14-laion2B-s32B-b79K which we have been using until that point. We will leave that swap for future iterations and will also consider using higher resolution images.”
* Datasets
* Together’s RedPajama-Data-v2: An open dataset with 30 trillion tokens for training large language models
* LAION COCO: 600M synthetic captions from Laion2B-en
* Chip Huyen’s writeup on LMMs
* Joseph Nelson of Roboflow on Latent Space
* HuggingFace M4
* HuggingFace timm: library containing SOTA computer vision models, layers, utilities, optimizers, schedulers, data-loaders, augmentations, and training/evaluation scripts. It comes packaged with >700 pretrained models, and is designed to be flexible and easy to use.
* Logan Kilpatrick declaring 2024 the year of Multimodal AI at AI Engineer Summit
This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe
SaaStr 717: How Enterprise SaaS Companies are Buying AI (or Not) with ContextualAI, Anthropic, and Glean
SaaStr 717: How Enterprise SaaS Companies are Buying AI (or Not) with ContextualAI, Anthropic, and Glean
While AI seems to have gone mainstream for consumers and smaller SaaS companies, but what about the big guys?
While the first generation of Generative AI is great, it's not quite ready to solve Enterprise problems. So, where are we now in the adoption cycle for the Enterprise world?
We brought together an Enterprise SaaS panel with:
Douwe Kiela, CEO of ContextualAI
Benjamin Mann, co-founder of Anthropic
Arvind Jain, CEO of Glean
and Sandhya Hedge, General Partner at Unusual VC
To help us figure out how to sell GenAI software to some of the biggest organizations in the world.
--------------------------------------------------------------------------------------------
SaaStr hosts the largest SaaS community events on the planet.
Join us in 2024 at:
SaaStr Annual: Sept. 10-12 in the SF Bay Area. Join 12,500 SaaS professionals, CEOs, revenue leaders and investors for the world's LARGEST SaaS community event of the year. Podcast listeners can grab a discount on tickets here: https://www.saastrannual2024.com/buy-tickets?promo=fave20
SaaStr Europa: June 5-6 in London. We'll be hosting the 5th SaaStr Europa in London for two days of content and networking. Join 3,000 SaaS and Cloud leaders. Podcast listeners can grab a discount on Europa tickets here: https://www.saastreuropa2024.com/buy-tickets?promo=fave200
--------------
This episode is sponsored by: Northwest Registered Agent
When starting your business, it's important to use a service that will actually help you. Northwest Registered Agent is that service. They'll form your company fast, give you the documents you need to open a business bank account, and even provide you with mail scanning and a business address to keep your personal privacy intact. Visit https://www.northwestregisteredagent.com/saastr to get a 60 percent discount on your next LLC.
AI is dangerous, but not for the reasons you think | Sasha Luccioni
AI won't kill us all — but that doesn't make it trustworthy. Instead of getting distracted by future existential risks, AI ethics researcher Sasha Luccioni thinks we need to focus on the technology's current negative impacts, like emitting carbon, infringing copyrights and spreading biased information. She offers practical solutions to regulate our AI-filled future — so it's inclusive and transparent.
Hosted on Acast. See acast.com/privacy for more information.
Smart Talks with IBM - Hugging Face and watsonx: Why Open Source Is the Future of AI in Business
Open-source innovation is the future of AI. In this episode of Smart Talks with IBM, Malcolm Gladwell and Tim Harford discuss the open-source AI community with Jeff Boudier, head of product and growth at Hugging Face. They chat about the history and future of open-source AI, its critical importance to AI progress, the IBM watsonx partnership with Hugging Face, and how businesses can leverage open-source AI for their specific needs.
Visit us at ibm.com/smarttalks
Learn more about the Hugging Face partnership: https://newsroom.ibm.com/2023-08-24-IBM-to-Participate-in-235M-Series-D-Funding-Round-of-Hugging-Face
This is a paid advertisement from IBM.
See omnystudio.com/listener for privacy information.
#430: Delightful Machine Learning Apps with Gradio
See the full show notes for this episode on the website at talkpython.fm/430
20VC: The Biggest AI Leaders on What Matters More; Model Size or Data Size & Where Does The Value in AI Accrue; to Startups or to Incumbents
Richard Socher is the founder and CEO of You.com. Richard previously served as the Chief Scientist and EVP at Salesforce.
Douwe Kiela is the CEO of Contextual AI, building the contextual language model to power the future of businesses. Previously, he was the Head of Research at Hugging Face, and before that a Research Scientist at Facebook AI Research.
Alex Lebrun is the Co-Founder and CEO of Nabla, an AI assistant for doctors. Prior to Nabla, he led engineering at Facebook AI Research. Alex founded Wit.ai, acquired by Facebook in 2015.
Tomasz Tunguz is the Founder and General Partner @ Theory Ventures, just announced last week, Theory is a $230M fund that invests $1-25m in early-stage companies that leverage technology discontinuities into go-to-market advantages.
Sarah Guo is the Founding Partner @ Conviction Capital, a $100M first fund purpose-built to serve "Software 3.0" companies. Prior to founding Conviction, Sarah was a General Partner at Greylock where she made investments in the likes of Figma, Coda and Neeva.
Emad Mostaque is the Co-Founder and CEO @ StabilityAI, the parent company of Stable Diffusion. Stability are building the foundation to activate humanity's potential. To date, Emad has raised over $110M with Stability with the latest round reportedly pricing the company at $4BN.
Clem Delangue is the Co-Founder and CEO @ Hugging Face, the AI community building the future. To date, Clem has raised over $160M from the likes of Sequoia, Coatue, Addition and Lux Capital to name a few.
Cris Valenzuela is the CEO and co-founder of Runway, the company that trains and builds generative AI models for content creation. To date, Cris has raised over $285M for the company from the likes of Lux Capital, Felicis, Coatue, Amplify, and Nvidia to name a few.
Noam Shazeer is the co-founder and CEO of Character.AI. A renowned computer scientist and researcher, Shazeer is one of the foremost experts in artificial intelligence (AI) and natural language processing (NLP).
The Two Most Pressing Questions in AI:
What matters more the size of the model or the size of the data?
Where does the value accrue in the next 5-10 years; to startups or to incumbents?
695: NLP with Transformers, feat. Hugging Face's Lewis Tunstall
What are transformers in AI, and how do they help developers to run LLMs efficiently and accurately? This is a key question in this week’s episode, where Hugging Face’s ML Engineer Lewis Tunstall sits down with host Jon Krohn to discuss encoders and decoders, and the importance of continuing to foster democratic environments like GitHub for creating open-source models.
This episode is brought to you by the AWS Insiders Podcast, by WithFeeling.ai, the company bringing humanity into AI, and by Modelbit, for deploying models in seconds. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.
In this episode you will learn:
• What a transformer is, and why it is so important for NLP [04:34]
• Different types of transformers and how they vary [11:39]
• Why it’s necessary to know how a transformer works [31:52]
• Hugging Face’s role in the application of transformers [57:10]
• Lewis Tunstall’s experience of working at Hugging Face [1:02:08]
• How and where to start with Hugging Face libraries [1:18:27]
• The necessity to democratize ML models in the future [1:25:25]
Additional materials: www.superdatascience.com/695
20VC: Why Data Size Matters More Than Model Size, Why The Google Employee Was Wrong; OpenAI and Google Have the Advantage & Why Open Source is Not Going to Win with Douwe Kiela, Co-Founder @ Contextual AI
Douwe Kiela is the CEO of Contextual AI, building the contextual language model to power the future of businesses. Last month Contextual closed a $20M funding round including Bain Capital, Sarah Guo, Elad Gil and 20VC. He is also an Adjunct Professor in Symbolic Systems at Stanford University. Previously, he was the Head of Research at Hugging Face, and before that a Research Scientist at Facebook AI Research.
In Today's Episode with Douwe Kiela We Discuss:
1. Founding a Foundational Model Company in 2023:
How did Douwe make his way into the world of AI and ML over a decade ago?
What are some of his biggest lessons from his time working with Yann LeCun and Meta?
How does Douwe's background in philosophy help him in AI today?
2. Foundational Model Providers: Challenges and Alternatives:
What are the biggest problems with the existing foundational data models?
Will there be one to rule them all? How does the landscape play out?
Why does Douwe believe OpenAI's data acquisition strategy has been the best?
3. Data Models: Size and Structure:
Why does Douwe believe it is naive to think the open approach will beat the closed approach?
What are the biggest downsides to the open approach?
Does the size of data model matter today? What matters more?
How important is access to proprietary data? Are VCs naive to turn down founders due to a lack of access to proprietary data?
4. Regulation and the World Around Us:
How does Douwe expect the regulatory landscape to play out around AI?
Why is Europe the worst when it comes to regulation? Will this be different this time?
How does Douwe analyse Elon's petition to pause the development of AI for 6 months?
Do founders building AI companies have to be in the valley?
20VC: Who Wins the AI Race; Startups or Incumbents & Does Having Proprietary Data Really Matter For Startups Today?
One of the core questions in AI and investing today; who wins, startups or incumbents? Startups have speed and innovation but incumbents have scale, resources, and distribution?
Today we hear from 6 leading investors and founders discussing where they place their bets who has the advantage; startups or incumbents?
Emad Mostaque is CEO @ StabilityAI, the parent company of Stable Diffusion. To date, Emad has raised over $110M with Stability with the latest round reportedly pricing the company at $4BN.
Yann LeCun is VP & Chief AI Scientist at Meta and Professor at NYU. He was the founding Director of FAIR and of the NYU Center for Data Science.
Clem Delangue is the Co-Founder and CEO @ Hugging Face, the AI community building the future. Clem has raised over $160M from the likes of Sequoia, Coatue, Addition and Lux Capital to name a few.
Sarah Guo is the Founding Partner @ Conviction Capital, a $100M first fund purpose-built to serve "Software 3.0" companies. Prior to founding Conviction, Sarah was a General Partner at Greylock.
Vince Hankes is a Partner @ Thrive Capital where he has led the firm's investments in OpenAI, Melio, and Airplane.dev. Prior to Thrive, Vince learned the craft of venture from Lee Fixel @ Tiger.
Tomasz Tunguz is the Founder and General Partner @ Theory Ventures, a $230M fund that invests $1-25m in companies that leverage technology discontinuities into go-to-market advantages.
The Question of the Day:
Who wins? Startups or Incumbents?
20VC: Why The Future of AI Is Open Not Closed, Why We Are Years Away From AI Being Autonomous, Why AI Founders Do Not Need to Move to the Valley & Why Founders Should Not Meet Investors in Between Rounds with Clem Delangue @ Hugging Face
Clem Delangue is the Co-Founder and CEO @ Hugging Face, the AI community building the future. To date, Clem has raised over $160M from the likes of Sequoia, Coatue, Addition and Lux Capital to name a few. Prior to Hugging Face, Clem was in product and marketing at two different startups both of which were acquired.
In Today's Episode with Clem Delangue:
1. From Tamagotchi to Leading the World of AI:
How did a Tamagotchi startup turn into one of the hottest AI startups in the world?
What does Clem know now that he wishes he had known when he started?
What are Clem's biggest pieces of advice to founders on pivoting?
2. AI: Trend or Transformation:
To what extent does Clem believe the current hype in AI is justified?
What is overblown? What have been some true and groundbreaking developments?
How far away does Clem believe AGI is?
What is a massive misconception the public has that Clem wishes he could change?
3. Open vs Closed: Which Model Wins:
Why does Clem believe the future of AI will be won by open-source?
What is his reasoning to suggest closed is fundamentally a weaker model?
Does Clem acknowledge that in the short term, enterprises will buy from a closed model with greater ease? How does he plan to tackle this?
4. Regulation: What Happens Now:
What regulatory changes need to be made in the world of AI most urgently?
Is Elon Musk right to suggest the immediate pausing of developments in AI?
What does Clem believe to be the most likely scenario to AI regulation in the next 12 months?
5. Fundraising: Lessons and Reflection on Raising $160M:
Do AI startups fundamentally cost more money than normal startups to build?
Why does Clem not meet investors in between rounds?
What does Clem believe is the most helpful thing an investor can do?
What are Clem's spiciest takes on venture as a financing model?
Open Source Generative AI at Hugging Face with Jeff Boudier - #624
Today we’re joined by Jeff Boudier, head of product at Hugging Face 🤗. In our conversation with Jeff, we explore the current landscape of open-source machine learning tools and models, the recent shift towards consumer-focused releases, and the importance of making ML tools accessible. We also discuss the growth of the Hugging Face Hub, which currently hosts over 150k models, and how formalizing their collaboration with AWS will help drive the adoption of open-source models in the enterprise.
The complete show notes for this episode can be found at twimlai.com/go/624
Why the Future of Machine Learning is Open Source with Huggingface’s Clem Delangue
After starting as a talking emoji companion, Hugging Face is now an organizing force for the open source AI research ecosystem. Its models are used by companies such as Apple, Salesforce and Microsoft, and it's working to become the GitHub for ML.
This week on the podcast, Sarah Guo and Elad Gil talk to Clem Delangue, co-founder and CEO of Hugging Face. Clem shares how they shifted away from their original product, why every employee at Hugging Face is responsible for community-building, the modalities he's most interested in, and what role open source has in the AI race.
Show Links:
Hugging Face website
The $2 Billion Emoji: Hugging Face Wants To Be Launchpad For A Machine Learning Revolution - Forbes
Sign up for new podcasts every week. Email feedback to show@no-priors.com
Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @ClementDelangue
Show Notes:
[01:53] - how Clem first became interested in ML, being shouted at by eBay sellers, and the foretelling of the end of barcode scanning
[3:34] - early iterations of Hugging Face, trying to make a less boring AI tamagotchi, and switching directions towards open source tools
[5:36] - advice for founders considering a change in direction, 30%+ experimentation
[7:39] - 1st users, MLTwitter, approach to community
[10:47] - enterprise ML maturity, days to production
[12:54] - open source vs. proprietary models
[15:56] - main model tasks, architectures and sizes
[19:12] - decentralized infrastructure, data opt out
[24:16] - Hugging Face’s business model, GitHub
[28:09] - What Clem is excited about in AI
Multimodal, Multi-Lingual NLP at Hugging Face with John Bohannon and Douwe Kiela - #589
In this extra special episode of the TWIML AI Podcast, a friend of the show John Bohannon leads a jam-packed conversation with Hugging Face’s recently appointed head of research Douwe Kiela. In our conversation with Douwe, we explore his role at the company, how his perception of Hugging Face has changed since joining, and what research entails at the company. We discuss the emergence of the transformer model and the emergence of BERT-ology, the recent shift to solving more multimodal problems, the importance of this subfield as one of the “Grand Directions'' of Hugging Face’s research agenda, and the importance of BLOOM, the open-access Multilingual Language Model that was the output of the BigScience project. Finally, we get into how Douwe’s background in philosophy shapes his view of current projects, as well as his projections for the future of NLP and multimodal ML.
The complete show notes for this episode can be found at twimlai.com/go/589
🤗 The AI community building the future
Hugging Face is increasingly becomes the “hub” of AI innovation. In this episode, Merve Noyan joins us to dive into this hub in more detail. We discuss automation around model cards, reproducibility, and the new community features. If you are wanting to engage with the wider AI community, this is the show for you!
Featuring:
Merve Noyan – GitHub, X
Chris Benson – Website, GitHub, LinkedIn, X
Daniel Whitenack – Website, GitHub, X
Show Notes:
The Hugging Face hub
New Hugging Face community features
Upcoming Events:
Register for upcoming webinars here!
.jpg?width=400){kind=small}