Julien Chaumond
@julien_c
Followers
57K
Following
23K
Media
2K
Statuses
16K
Co-founder and CTO at @huggingface 🤗. ML/AI for everyone, building products to propel communities fwd. @Stanford + @Polytechnique
Paris / New York
Joined September 2008
Some people said that closed APIs were winning. but we will never give up the fight for open source AI ⚔️⚔️. Today is a big day as we launch the first open source alternative to ChatGPT:. HuggingChat 💬. Powered by Open Assistant's latest model – the best open source chat
189
995
4K
The good news is, . with the $6 billion he just raised,. Elon Musk could buy Twitter at current valuation.
99
86
2K
JUST IN: The board of OpenAI will quit if the OpenAI board can't provide evidence of why they fired Sam Altman.
63
81
2K
If only we already had a few cool AI companies in Paris.
45
107
2K
meanwhile I’m still on free @GoogleColab.
First @NVIDIA DGX H200 in the world, hand-delivered to OpenAI and dedicated by Jensen "to advance AI, computing, and humanity":
43
75
2K
Let’s make AI OPEN 🤯. Yesterday we've announced we @huggingface are partnering with @awscloud . Now, here's my raw CTO take:
27
217
2K
BTW. dalle-mega from @borisdayma is now openly accessible on @huggingface. ⚡️ To download it (10GB): git clone
24
183
1K
If I was starting a company today, I would look into productizing this model 👀.
Microsoft released a groundbreaking model that can be used for web automation, with MIT license 🔥👏. OmniParser is a state-of-the-art UI parsing/understanding model that outperforms GPT4V in parsing. 👏
16
88
1K
come and join us at HF ❗️.
31
17
1K
exactly.
15
52
1K
No idea tbh but one day we might find out.
I don’t get HuggingFace - how is this even a business.
30
31
1K
Qwen QwQ switching to Chinese when it needs to _really think_ about something, then switching back to English, is pretty cool. @Alibaba_Qwen
50
102
1K
i love the hf team so much. (but i don't even need a coup to say it).
38
67
984
🦄 We ported @openai's GPT-2 to run on-device (using Swift and CoreML on iOS) 🔥🔥. Large transformers models can now live on the edge. 📱📲. The video below is GPT-2 running locally (no network) on the device!. Code: Built w/ @LysandreJik at @huggingface
14
200
921
Prediction:. there'll be 10 open reproductions of ChatGPT in 6 months. and if i understand correctly, models are quite smaller than LLMs so they might be significantly easier to deploy at scale (maybe even single-GPU).
57
84
887
🔥 Introducing Tokenizers: ultra-fast, extensible tokenization for state-of-the-art NLP 🔥. ➡️
10
201
873
we just shipped HuggingChat on iOS 💬. The app is super polished and gives you access to the community's best open AI models, on the go. Give it a try!. link to Appstore below ⤵️
76
140
853
is it me or has ML Twitter been way way quieter recently?.
99
26
800
Ten days ago I posted about GPT Store being a bit sad 😢:. What if we could build an open source alternative, with the full power of the Community?. So last Friday we launched Hugging Chat Assistants, and the adoption has been impressive:.- 4,000 Assistants have been created on
39
128
773
This changes everything. HuggingChat with Web Search is now available. It fetches Google's top results to help the model generate its answer using data from the Web. And it's free 🤯
22
128
749
Chat-UI, the codebase behind our HuggingChat app, is now open sourced on GitHub. 🎉. We are looking for contributors 🏴☠️. Do YOU want to push the state of Open source AI forward? ☝️. The codebase is in @sveltejs SvelteKit but you can contribute in many different ways, not just
10
143
631
🔥🔥 Series A!! 🔥🔥. Solving Natural language is going to be the biggest achievement of our lifetime, and is the best proxy for Artificial intelligence. Not one company, even the Tech Titans, will be able to do it by itself – the only way we'll achieve this is working together
15
90
635
At Hugging Face we have crazy engineers like @MorganFunto who are building racks of M1 Mac minis. The goal is to try and build the most efficient FLOPS/$ machine learning rig. AND the most energy-friendly CO2_eq/FLOPS machine learning rig. Has anyone here tried something like it?
47
47
614
Having a pretty rough week. Please consider sending hugs! 🤗🤗.
210
2
612
Apple, are you trying to bankrupt us?. At ~500 MB per average model download, 90k hits (Apple's IP addresses are the 17. block) translate to ~45TB of downstream bandwidth… which starts being costly for us. Your new ML platform Overton ( ⤵️
27
126
598
GPT-2 on device is blazing fast on iPhone 11 ⚡️. Core ML 3 is officially out so we can do state-of-the-art text generation on mobile (117M parameters running ~3 times par second on the neural engine!). We put together a small video benchmark ⬇️
9
153
603
We have decided to update text-generation-inference (TGI)'s license. We switch the license from HFOIL (our custom license) back to Apache 2, hence making the library fully open-source. Read below for why we are making this change 👀
23
100
569
my prediction for 2024 (yes, i have only one) 💡. Local ML is going to be huge. It will be in part driven by the adoption of Apple Silicon and other innovative hardware, but also on raw CPU and mobile devices. In many cases except for the largest of LLMs, local inference will.
48
75
549
Ok who’s in to pool some resources with us and replicate GPT-3 in the open?.
45
59
544
Genuinely wondering whether Prompt engineering will still be a thing at the end of 2023.
109
28
523
Today we’re announcing 3 big things on the @huggingface Hub 🔥. Open this 🧵 to see all 3️⃣ of them. I'm very excited ❤️. 1️⃣ The first one is that we’ve just rolled out Spaces GPU Upgrades for everyone. You can now upgrade to T4 and A10G, and we have A100 in private beta.
6
102
514
Make a machine learning engineer happy in 1 command line: 👀.
148
28
465
Please let our engineers sleep!! 🙏.
Introducing: Zephyr 141B-A35B 🥁. 🔥Mixtral-8x22B fine-tune.🤯 Using DORPO: new alignment algorithm (no SFT, open ).🚀 With 7k instances of (open) data. Very strong IFEval, BBH, AGIEval. Enjoy! 🤗.
10
21
449
Introducing. HF Training Cluster as a service! 🔥🔥. Access to a large compute cluster is key for large-scale model training, but historically it's been hard to secure access to large numbers of hardware accelerators, even with a hefty budget. 💰. With Training cluster as a
17
92
439
Very happy to announce that we've moved our @huggingface AutoTrain compute to Quebec 🇨🇦(running on @ovhcloud). Our compute now emit 11x less CO₂ 🔥 and runs on 99% renewable energy🍀. Hat tip to our Infra team, to @electricityMap for the very neat CO₂ maps, and to @SashaMTL 💚
11
52
404
sending many hugs to the @huggingface Infra Team who stayed up all night to mitigate a db scaling issue. Post mortem will follow later in the week, in the meantime please join me in thanking Remy/Michelle/Eliott/Adrien/. you folks all rock 💎
22
23
409
Congrats to @allen_ai for the excellent release of OLMo. It's a true open source end-to-end release:. not just.• model code.• model weights. but also.• training code.• training data (and associated toolkits).• eval toolkits. This is pushing the enveloppe of open-source AI 🔥
6
84
404
What's the AI equivalent for forking a software repo?. Finetuning a model!. Today we introduce a new feature on the Hub:. We now display which base model a given model is:.• finetuned from,.• or is an adapter (LoRA, PETs, etc) for. This also enables a genealogic graph of all
13
54
408
I'm beyond stoked to launch the v2 of the @huggingface model hub today 🔥. Each of our 2,000 models now has an inference widget that lets you try it (text-classification, token-classification, translation, etc.) directly from the model page. It's all powered by the community 💖
5
99
400
We just helped a team member from Mongolia get a work visa to France. June 10: offer sent to candidate(&accepted).June 16: appointment at the FR consulate in Ulaanbaatar, Mongolia 🇲🇳.June 22: visa application approved.June 30: @mishig25 will travel to France 🎉. Go France!! 🇫🇷🇫🇷
13
46
389
By the way, we just updated 𝗢𝗨𝗥 𝗗𝗘𝗙𝗜𝗡𝗜𝗧𝗜𝗩𝗘 𝗧𝗨𝗧𝗢𝗥𝗜𝗔𝗟 🔥 to training a new language model from scratch. Updated Colab notebook with the new Trainer =). ➡️
3
88
384
The famous @ycombinator dropped a very cool article to showcase their 25+ companies who have trained their own AI models, rather than just use someone else's closed model through an API like a black box. ⤵️
7
46
388
Update to 𝗢𝗨𝗥 𝗗𝗘𝗙𝗜𝗡𝗜𝗧𝗜𝗩𝗘 𝗧𝗨𝗧𝗢𝗥𝗜𝗔𝗟 🔥. You can now click the "Open in Colab" link and run as a simple notebook. Hat/tip @srush_nlp for the idea and @AdityaMalte for the implem help 🔥. ➡️
6
81
373
Llama 2 (70B) just landed in HuggingChat💬. This is the largest running version of the model from @MetaAI, running on fast optimized inference on @huggingface infra. Unleash the llamas! 🦙🦙. Try it out now.
10
87
370
llama2-7b in the browser – using WebNN – is going to be 🔥🔥. on-device, local ML 💪 cc @xenovacom
5
49
365
Outdated we’re at 200x now!! 🚀. 😂😂.
Generative AI company revenue multiples. • @OpenAI 18.1x.• @AnthropicAI 20.5x.• @AI21Labs 28x.• @glean 56.4x.• MosaicML 65x (acq by Databricks).• @perplexity_ai 65x.• @harvey__ai 71.5x.• @cohere 100x.• @huggingface 150x . OpenAI a bargain? :).
14
24
359
personally a bit bored by 530 billion parameters models . I'd be more interested in a model that can be used or researched by 530 million humans.
16
31
349
I am super excited to announce that we've acquired XetHub! 🎊. The @xetdata team has developed technologies to enable Git to scale to TeraByte-size repositories. Under the hood they've been adding file chunking and deduplication inside Git. This will help us unlock the next 5
17
62
353
i rarely subtweet, but: do you actually need $580 million to build the future of AI? 🤔.
58
14
330
✨ Distil-All The Things✨. Today we release:.- 🐎DistilGPT-2 support in Write With Transformer. Compare your GPT-2 generations to our distilled, faster model: - 🚨 DistilBERT for question answering for iOS.- 🦄 DistilGPT-2 for iOS:
4
88
343
GPT-4 quality seems to be going down lately. Will you join the Empire, or the rebellion?
28
33
329
NVIDIA A100 GPUs are now available on @huggingface Spaces, at those competitive prices= 🔥🔥🔥. • USD 4.13/hour in self-serve.• lower prices for Enterprise customers. Spaces will expose more cool hardware for ML in the coming months, ping us if you have specific needs 🙏
7
45
333
i have a confession to make. not once in my life have I used pandas.
65
1
323
🔥 Thrilled to release our Swift Core ML implementation of BERT for question answering.🔥🔥. Transformers models now also live on the edge. 📱📲. You now CAN do state-of-the-art NLP on mobile devices!. Built w/ @LysandreJik and @Thom_Wolf at @huggingface
5
93
324
Can we say that all researchers at FAIR are now doing Meta learning?.
9
19
314
Nice Medium on how to serve @huggingface BERT in production with pytorch/serve by MFreidank . Hat/tip @joespeez.
1
97
316
if i didn't already have a startup, i'd be looking at right now and thinking about what to build on top of @AdeptAILabs's super cool model.
11
46
308
interview question:. to move 1 EB (1 ExaByte) of data, how many truck trailers of hard drives do you need?.
130
12
308
The @ollama - @huggingface integration has been rolled out for 1 week now, how it’s going?. Obviously, pretty well! We’re having on average 4500 pulls per day. That’s about one pull every 20 seconds!. What’s the top models you may ask?. Llama-3.2-1B still on top thanks to its
9
37
306
Luckily, they never invite CTO’s.
Why do CEOs have to go to in-person events like Davos or the Trump inauguration? I prefer to stay at the office and get things done haha.
6
4
305
@huggingface @awscloud This is not about BLOOM or ChatGPT. This is about the dozens of BLOOMs and ChatGPTs that are going to be released by the community in the coming months, and years.
6
25
290
For the record:. we've always been pretty frugal/scrappy at @huggingface. This is the first >$100k piece of equipment we've bought, ever. I still believe ideas are worth way more than compute (though we do have access to ample compute now). Photo credit: @gloriamika @MorganFunto
27
8
289
GPT Store is kinda sad TBH, i wish the Open Source AI community would build something way better 👋.
42
17
277
Since time immemorial,. the number 1 rule for engagement on Twitter/X has been,. to mention elonmusk, ylecun, or both.
20
11
283
10. Billion? 😬. I guess maybe 10 million was not impressive enough 💰
32
12
283
After some heated discussion 😬, we clarify our intent re. storage limits on the @huggingface Hub. TL;DR:. - public storage is free, and (unless blatant abuse) unlimited. We do ask that you consider upgrading to PRO and/or Enterprise Hub if possible. - private storage is paid
26
50
291
Introducing CamemBERT in @huggingface transformers!. New state-of-the-art in French NLU🇫🇷. Thanks @louismrt for contributing the implem 🙏. Thanks @ben_mlr, @pjox13, Yoann Dupont, Laurent Romary, Eric Villemonte de la Clergerie, @zehavoc, @bensagot @InriaParisNLP and @fb_research
9
61
276
I'm gonna memorize commit hash `e02037b3524686b57c5a861ea49ac751f15568af` forever ❤️❤️🔥.
4
5
280
Breaking news!. You can now browse from a @huggingface model page, to its: .- fine-tunes.- adapters.- merges.- quantized versions. and browse through the models' genealogy tree 🌲
11
49
285
We just passed 1️⃣0️⃣,0️⃣0️⃣0️⃣ models on @huggingface 🔥. This amounts to more than 1 trillion (10^12) machine learned parameters 🤯. Here's to the next 100k models 💜. Here's to the community building the future of AI with us ❤️🧡🖤
3
40
278
BERT is pretty safe these days because when he leaves his house he can always use a [MASK].
@julien_c pretty sure universal basic mask is like a new type of language model or something.
5
48
273
You can now filter for models that include data about their training's energy consumption and equivalent CO2 emissions ⚡️. To add this metadata to your models, see the screenshot below. hat/tip @SashaMTL @abhi1thakur @mmitchell_ai 🙏
14
44
266