Hugging Face, every mention
114 scenes, the whole family · ← back to Hugging Face
tap a year for its mentions
every year anyone Shawn Wang 24Nathan Lambert 19Jeremy Howard 11Swyx (Marcos Swix) 6Mitesh Agrawal 5Ethan Sutin 5Alessio Fanelli 5Vasek Mlejnsky 4Omar Sanseviero 4Michael Royzen 4
Verbatim, from the transcripts: the passages where Hugging Face comes up
Orbs: Shifting Coding to Cloud — Quinn Slack, Amp Code
- ▶ 17:31 Quinn Slack And we've seen that agents are very good at escaping containment with the whole like hugging face to debacle and all of that.
Next 100x in AI: Inference, Networking, & Self-Optimizing Models — Philip Kiely & Ali Taha, Baseten
- ▶ 12:53 Shawn Wang Ok, so, like, you know, a lot of people, all you guys, right, whenever a new model launch, like, people rush to say, like, oh, Hugging Face supports this, Fireworks supports this, Base 10 supports this, and I'm like, yeah, of course you…
- ▶ 1:32:58 Philip Kiely One big part of my job a couple years ago was for any arbitrary model that came out on Hugging Face, writing a config for it and kind of getting it up and running, and now the get it up and running config is, is one-shot-able, um, and so,…
The Future of AI Infra: from Kubernetes to Agent Sandboxes — Akshat Bubna, Modal CTO
- ▶ 12:26 unnamed speaker They'll take a hugging face link, and, you know, like, there's so much value just right there, right?
🔬 The Limits of AI in Science - Why We Need Self-Driving Labs — Joseph Krause, Radical AI
- ▶ 1:10:26 unnamed speaker The, the model and the benchmark are available on Hugging Face.
⚡️ Google's Open AI Strategy — Omar Sanseviero, Google DeepMind
- ▶ 3:42 Omar Sanseviero So for example, we work with Lama CPP, Olama, MLX, Hogan Faces, BLM, NVIDIA, AMD.
- ▶ 25:21 Omar Sanseviero Colab with Transformers or Oncelot or whichever library of your choice.
- ▶ 25:31 Omar Sanseviero Like Hogan face has skills, like all of these libraries have skills.
- ▶ 26:29 Omar Sanseviero I think maybe three, four years ago, we, we did a, like a nice interview about how I was growing, like, uh, at FoggingFace and how we were thinking, like, DevRel should look like.
⚡️ Reverse Engineering OpenAI's Training Data — Pratyush Maini, Datology
- ▶ 21:42 Pratyush Maini And then much faster than anything that hugging face or pajama does.
- ▶ 26:24 unnamed speaker and then it's like Microsoft or whatever, and it's Apple, then it's Hugging Face, then it's NVIDIA, now it's you guys, and I'm like, you know, where, where's the, the, the sort of persistence, or like, is this such a competitive field?
Captaining IMO Gold, Deep Think, On-Policy RL, Feeling the AGI in Singapore — Yi Tay
- ▶ 35:49 unnamed speaker Like on Hug Your Face?
Artificial Analysis: The Independent LLM Analysis House — with George Cameron and Micah Hill-Smith
- ▶ 29:40 Shawn Wang You know, we have done a few evals podcasts over the, over the years, and we did one with Clementine of Hugging Face, who maintains the open source leaderboard.
- ▶ 53:55 Shawn Wang It's hugging face. 3 times in the scene
[State of AI Papers 2025] Fixing Research with Social Signals, OCR & Implementation — Team AlphaXiv
- ▶ 3:08 unnamed speaker I guess the other, the, the reason that when I first saw AlphaKive, I was, I didn't necessarily think about it as new, was because I knew that Hugging Face had also launched, like, a paper discussion thing, and like, well, 3 times in the scene
- ▶ 6:12 unnamed speaker So, I think for the first year or two of working on it, it was a project, and what really forced that transition for us was we were looking at, you know, papers with code, weights and biases, hugging face, and we're like, okay, we have a, 2 times in the scene
[State of Evals] LMArena's $1.7B Vision — Anastasios Angelopoulos, LMArena
- ▶ 8:52 Shawn Wang Did you tell the Hugging Face folks that? 4 times in the scene
SAM 3: The Eyes for AI — Nikhila & Pengchuan (Meta Superintelligence), ft. Joseph Nelson (Roboflow)
- ▶ 19:31 unnamed speaker I would say Hugging Face has been doing a lot here, uh, on, uh, and, and other, other companies.
⚡ Open Model Pretraining Masterclass — Elie Bakouch, HuggingFace SmolLM 3, FineWeb, FinePDF
- ▶ 0:33 Shawn Wang Model training, all that stuff with Ellie from Hugging Face. 3 times in the scene
- ▶ 4:24 Elie Bakouch I didn't say that was my job at hearing face is being the guy that's like keeping touch with what's like the, the pre-training, uh, innovation are.
- ▶ 28:40 Elie Bakouch A GPT OSS coin three doesn't have Shard Expert, but like the coin three next, which is like, uh, really is like, uh, I mean, they didn't release it yet, but they submit the PR to Transformers.
- ▶ 34:59 Elie Bakouch For example, we tried, uh, we are training, uh, MOE currently at TargetFace.
- ▶ 48:51 Elie Bakouch Um, and you know, there is this paper from, um, actually from people at HuginFace at the time that is saying that you, you basically can repeat your data, uh,
- ▶ 54:39 Shawn Wang LightEval, I mean, I think a lot of people are like trying to move from the hugging face harness and, uh, 3 times in the scene
- ▶ 1:00:13 Alessio Fanelli Um, I found that a little clunky right now, just to find something that is token based because the hugging face API charges by the hour, which. 2 times in the scene
⚡️ Beyond Transformers with Power Retention
- ▶ 15:40 Diego Bachman This was released by the big code projects or hugging face affiliated.
- ▶ 24:13 Diego Bachman So just like right now, there's a large ecosystem of open source transformers on hugging face and other places that lots of people use and benefit from.
Context Engineering for Agents - Lance Martin, LangChain
- ▶ 28:17 Lance Martin HuggingFace actually has a very interesting OpenDeep Research implementation.
A Technical History of Generative Media
- ▶ 4:23 Batuhan Taskaya Twitter, Reddit, uh, you know, Hugging Face, seeing how popular the models are in Hugging Face, uh, and other demos.
- ▶ 38:04 Gorkem Yurtseven When you look at Hugging Face, let's look right now, I'm sure the top models are image models.
- ▶ 1:02:20 Batuhan Taskaya Like one, one of our, uh, one of our applied ML engineers is like, had the number one top hugging face space with like, you know, creative workflows, whatever. 2 times in the scene
Better Data is All You Need — Ari Morcos, Datology
- ▶ 33:03 Ari Morcos Um, so you're left to kind of, you know, the Allen Institute, things like DCLM, hugging face, et cetera, to make progress there.
⚡️Accelerators @ 3x NVIDIA H200 perf, Made in the USA - Thomas Sohmers + Mitesh Agrawal, Positron AI
- ▶ 18:58 Mitesh Agrawal We basically take the raw binary weights output from when you train your model on an NVIDIA GPU, you get a .pd.safetensor file that you upload on Hugging Face, and that's your raw binary weights for your model. 3 times in the scene
- ▶ 21:22 Thomas Sohmers And we were betting on the Hugging Face Transformer library and sort of the framework that, that was created there.
- ▶ 46:27 Mitesh Agrawal Bet on the Hugging Fish Transformers.
- ▶ 46:27 Mitesh Agrawal Bet on the Hugging Fish Transformers.
The RLVR Revolution — with Nathan Lambert (AI2, Interconnects.ai)
- ▶ 3:15 Nathan Lambert The academic community had been using this one data set since like all the way back in the hugging face models of like Zephyr beta is when this ultra feedback data set got popular.
- ▶ 56:57 Nathan Lambert And it'll be like, I think one of the early examples we had playing with this at Hugging Face was the model would just say JavaScript would be JavaScript, JavaScript.
- ▶ 1:06:49 Nathan Lambert Um, but then the last thing is for people doing research, it's like wacky model routing things where you figure out like a bunch of different models to off hugging face to route to, because an open model tool thing could use way more… 6 times in the scene
⚡️The Future of Notebooks - with Akshay Agrawal of Marimo
- ▶ 2:14 Akshay Agrawal And we're used at companies like OpenAI, Hugging Face, Cloudflare, BlackRock, universities like Stanford and Berkeley.
Information Theory for Language Models: Jack Morris
- ▶ 50:12 Shawn Wang For a given definition of open source, which is like, we released a waste of hugging your face. 2 times in the scene
The AI Coding Factory
Voice AI Masterclass — Kwindla Hultman Kramer and swyx
- ▶ 9:38 Kwindla Hultman Kramer Like Freddie from hugging face is going to talk about fast RTC and hang out in the discord and kind of be around.
Why Every Agent needs Open Source Cloud Sandboxes
- ▶ 23:23 Vasek Mlejnsky There's like a fun story from Hugging Face when they are using, when they were using us and are using us for their, uh, Open R one model.
- ▶ 55:04 Vasek Mlejnsky The way HuggingFace, who built, um, the OpenROne project is using us is during like the reinforcement learn, code gen reinforcement learning step where the ROne model, uh, the OpenROne model has a training step where they give it a 3 times in the scene
GPT 4.1: The New OpenAI Workhorse
- ▶ 11:20 Shawn Wang Uh, so I actually went into your hugging face release and got an example of the graph, uh, task.
The Magic of LLM Distillation — Rishabh Agarwal, Google DeepMind
- ▶ 32:52 Rishabh Agarwal I'm just saying the reason it's interesting is because, I mean, banking on any existing RL framework, this is how the Hugging Face people implemented this, but I'm saying that usually infra is a big bottleneck for these kind of approaches,…
Bee AI: The Wearable Ambient Agent
- ▶ 4:48 Ethan Sutin So it was just an app that you kind of chatted with and it would ask you questions and then like, give you some feedback, but Hugging Face first version was launched at the same time. 5 times in the scene
smol agents are all you need
- ▶ 0:11 Swyx (Marcos Swix) Hey, and today it's a very beautiful small episode because we have the creator of Small Agents, Emmerich from Hugging Face.
- ▶ 6:44 Swyx (Marcos Swix) I know, uh, Hugging Face has been working on small LM as well. 4 times in the scene
- ▶ 11:22 Swyx (Marcos Swix) Uh, but anyway, I think, you know, the, the French Hugging Face mafia, uh,
- ▶ 17:42 Aymeric (Emmerich) I think the way they went with this is really the way forward, and that's also why, that's the next step for us at Hugging Face.
Outlasting Noam Shazeer, Crowdsourcing Chai AI w/ 1.4m DAU — with William Beauchamp, Chai Research
- ▶ 11:44 Alessio Fanelli Because, you know, if you think about, if somebody told you that was the hugging face founding story, people might believe it.
- ▶ 47:29 William Beauchamp Ok, so what we built was we built Triverse, and Triverse is kind of, it's kind of like a prototype, is the way to think about it, and it started with this, this observation that, well, how many models get submitted to Hug and Face a day?
The State of Reasoning — from Nathan Lambert, Interconnects/AI2 [LS Live @ NeurIPS 2024]
- ▶ 11:48 Nathan Lambert Quickly, we'll see things like hugging base having more of these.
2024 Year in Review: The Big Scaling Debate, the Four Wars of AI, Top Themes and the Rise of Agents
- ▶ 1:55 Shawn Wang Hacker News hates it, which is a good sign, but there's enough people that have defined it, you know, GitHub, when you launch GitHub models, which is the Hugging Face clone, they put AI engineers at the, in the banner and like above the…
- ▶ 1:06:13 Shawn Wang There's stuff that's basically in V-one of the Hugging Face Open Models leaderboard, right?
Best of 2024: Synthetic Data / Smol Models, Loubna Ben Allal, HuggingFace [LS Live! @ NeurIPS 2024]
- ▶ 6:02 Loubna Ben Allal And as HuggingFace, because we're like open source, we tried to reproduce what they did.
- ▶ 18:19 Loubna Ben Allal For example, here, this is an app called Pocket Pal, where you can go and select a model from Hugging Face.
0 to over $8M ARR in 2 months as a Claude Wrapper (Bolt.new, Qodo)
- ▶ 46:23 Itamar Friedman And then there's, there's a clear, like, almost hugging face, like, model, like, yeah, you can do that, but why should you try to deploy it yourself, deploy it with us?
- ▶ 1:03:56 Eric Simons So think like, kind of like Google CoLab sort of thing, or like Hugging Face has their kind of version of this.
[Paper Club] BERT: Bidirectional Encoder Representations from Transformers
- ▶ 34:24 unnamed speaker So distilbert is a hugging face, uh, like recreation of Bert that, like, has very comparable performance on, uh, many fewer parameters. 4 times in the scene
- ▶ 35:05 unnamed speaker So a lot of this leans very heavily on the, uh, hugging face transformers library.
- ▶ 46:39 unnamed speaker Uh, no, the last time I looked at this, like, leaderboard and hugging face, uh, I think it was all led by these, uh, transformed LLMs, uh, that now get the best performance, like a Mistral seven B turned into, uh, embedding model.
Why Compound AI + Open Source will beat Closed AI — with Lin Qiao, CEO of Fireworks AI
- ▶ 13:06 Shawn Wang Basically, it's just like the meta version of whatever HuggingFace offers, you know, or TensorRT, or BLM, or whatever the open source opportunity is.
Agents @ Work: Lindy.ai (with live demo!)
- ▶ 57:44 unnamed speaker Hugging face.
[Paper Club] Upcycling Large Language Models into Mixture of Experts
- ▶ 12:37 Ethan He Uh, let's also look at the implementation of Mixtro eight by seven on Hagen-Phys transformer.
- ▶ 12:37 Ethan He Uh, let's also look at the implementation of Mixtro eight by seven on Hagen-Phys transformer.
- ▶ 15:13 Ethan He You can combine it with hugging phase transformer without any problem.
- ▶ 15:13 Ethan He You can combine it with hugging phase transformer without any problem.
[Paper Club] Molmo + Pixmo + Whisper 3 Turbo - with Vibhu Sapra, Nathan Lambert, Amgadoz
- ▶ 1:02:45 unnamed speaker The distal whisperer model has a good paper from the eigenphase team and goes into much more detail about this.
Production AI Engineering starts with Evals
- ▶ 17:56 Ankur Goyal I remember I was in New York, and I was playing with Burt on Hugging Face, which had made it, like, really easy at that point to actually do that, and I, they had, like, this little square, you know, in the, in the, in the, uh, right-hand… 2 times in the scene
- ▶ 33:17 Ankur Goyal I remember when I was first acclimating to this problem, I used, I had to learn how to use hugging face and weights and biases.
[Paper Club] Berkeley Function Calling Paper Club! — Sam Julien, Writer
- ▶ 16:42 Sam Julien They also posted the, the, the data set on hugging face, which is pretty interesting.
Answer.ai & AI Magic with Jeremy Howard
- ▶ 39:55 Jeremy Howard The hugging face library in peft doesn't really work in practice unless you use it with other things. 2 times in the scene
- ▶ 42:17 Jeremy Howard And even as we did so, new regressions were appearing in, like, Transformers and stuff, that Benjamin then had to go away and figure out, like, oh, how come flash attention doesn't work in this version of Transformers anymore with this set… 2 times in the scene
Segment Anything 2: Memory + Vision = Object Permanence — with Nikhila Ravi and Joseph Nelson
- ▶ 23:35 Shawn Wang Hugging face is probably already working on Transformers.js version of it, but totally makes sense.
The Winds of AI Winter (Q2 Four Wars of the AI Stack Recap)
- ▶ 8:31 unnamed speaker We talked about this with Clementine on the Hugging Face episode, and so we need to see what, what else, what is the next frontier?
- ▶ 40:38 unnamed speaker There's also related work from Hugging Face on the Numina math competition.
- ▶ 1:19:55 unnamed speaker Most of them, half of them come from the Hugging Face episode. 2 times in the scene
[LLM Paper Club] Llama 3.1 Paper: The Llama Family of Models
- ▶ 1:10:47 unnamed speaker Like, XLM is not deterministic at all, compared to maybe, I think, Transformers is more deterministic.
Training Llama 2, 3 & 4: The Path to Open Source AGI — with Thomas Scialom of Meta AI
- ▶ 17:03 Shawn Wang You know, I probably, it doesn't seem like it's your group, but you know, you also recently published mobile LLM, which on the small model side is, uh, is a really good research on just small model architecture that it looks like…
- ▶ 32:23 Alessio Fanelli From the Allen Institute who was at AgInFace leading RLHF before.
- ▶ 38:43 Alessio Fanelli So talking about evals, we just had an episode with Clementine from Hugging Face about leaderboards and arenas and evals and benchmarks and all of that.
- ▶ 47:59 Thomas Scialom And what I was looking right now on state-of-the-art results on Gaia, there's a leaderboard, by the way, you mentioned Clementine before, she contributed to Gaia as well, and Hugging Face put a leaderboard there on their website. 2 times in the scene
- ▶ 48:24 Thomas Scialom But OSCopilot then, and Autogen from Microsoft, and recently Hugging Face Agent, obtained some level one up to 60%.
The 10,000x Yolo Researcher Metagame — with Yi Tay of Reka
- ▶ 1:38:45 unnamed speaker Uh, Hugging Face recently did one, Datablations, which is like a data scaling loss paper, um, looking at data constraints, uh, which is, which is kind of nice.
- ▶ 2:01:23 Yi Tay But, but, uh, but now they're all gone, because people realize that, that you cannot climb LMSYS, because you need something more than just something that is lightweight, right, so I think that was just my, my overall, like, Honestly, the…
State of the Art: Training 70B LLMs on 10,000 H100 clusters
- ▶ 5:04 unnamed speaker And I think Hugging Face got it right.
Breaking down the OG GPT Paper by Alec Radford
- ▶ 45:32 unnamed speaker And this is actually how the model looks like if you try to load in the transformers framework.
A Comprehensive Overview of Large Language Models - Latent Space Paper Club
- ▶ 28:30 unnamed speaker So that's instruction fine tuning over here, um, and something called alignment tuning, where you want to ensure, um, that your model, uh, fulfills what people call the three H, the three H's of, uh, model behavior.
- ▶ 38:17 unnamed speaker So, um, essentially what happens is that if you go to maybe say, um, TensorFlow datasets or Hugging Face, you'll be able to download them, um, and then you'll be able to observe, um, these datasets, uh, by itself.
Open Source AI is AI we can Trust — with Soumith Chintala of Meta AI
- ▶ 1:13:01 Soumith Chintala Like, the blueprint here, I think, is you'd want someone to create a sinkhole for the feedback, some centralized sinkhole, like maybe hugging face or someone, uh, just funds, like, ok, like, I will make available a call to log a string…
A Brief History of the Open Source AI Hacker - with Ben Firshman of Replicate
- ▶ 52:39 unnamed speaker Mozilla came out with a llama file, and then, um, I don't know if this is in the same category even, but I'm just gonna throw it in there, like Hugging Face has the Transformers and Diffuses library, which is a way of disseminating models… 2 times in the scene
- ▶ 52:39 unnamed speaker Mozilla came out with a llama file, and then, um, I don't know if this is in the same category even, but I'm just gonna throw it in there, like Hugging Face has the Transformers and Diffuses library, which is a way of disseminating models… 3 times in the scene
The State of AI in production — with David Hsu of Retool
- ▶ 52:35 Shawn Wang Like I, we just released an episode today, uh, talking about IdaFix from HuggingFace.
The Four Wars of the AI Stack - Dec 2023 Recap
The Origin and Future of RLHF: the secret ingredient for ChatGPT - with Nathan Lambert
- ▶ 0:31 unnamed speaker Um, you were, you bootstrapped the RLHF team at Hugging Face, and you recently joined the Allen Institute as a research scientist.
- ▶ 41:21 Nathan Lambert We played with it a bit at Hugging Face.
page 1 of 2 · 100 scenes per page · newest episode first next →