Hugging Face's logo

huggingPartyParis

community
https://partiful.com/e/oWOMGoPxB5D37qw5F8yN
Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

TimDarcet  authored a paper 10 days ago
Cluster and Predict Latents Patches for Improved Masked Image Modeling
abenechehab  authored a paper 16 days ago
TAG: A Decentralized Framework for Multi-Agent Hierarchical Reinforcement Learning
abenechehab  authored a paper 23 days ago
Large Language Models Orchestrating Structured Reasoning Achieve Kaggle Grandmaster Level
View all activity

HuggingPartyParis's activity

eliebak 
posted an update about 22 hours ago
view post
Post
752
Google just dropped an exciting technical report for the brand-new Gemma3 model! 🚀 Here are my personal notes highlighting the most intriguing architectural innovations, design choices, and insights from this release:

1) Architecture choices:
> No more softcaping, replace by QK-Norm
> Both Pre AND Post Norm
> Wider MLP than Qwen2.5, ~ same depth
> SWA with 5:1 and 1024 (very small and cool ablation on the paper!)
> No MLA to save KV cache, SWA do the job!

2) Long context
> Only increase the rope in the global layer (to 1M)
> Confirmation that it's harder to do long context for smol models, no 128k for the 1B
> Pretrained with 32k context? seems very high
> No yarn nor llama3 like rope extension

3) Distillation
> Only keep te first 256 logits for the teacher
> Ablation on the teacher gap (tl;dr you need some "patience" to see that using a small teacher is better)
> On policy distillation yeahh (by
@agarwl_
et al), not sure if the teacher gap behave the same here, curious if someone have more info?

4) Others
> Checkpoint with QAT, that's very cool
> RL using improve version of BOND, WARM/WARP good excuse to look at
@ramealexandre
papers
> Only use Zero3, no TP/PP if i understand correctly ?
> Training budget relatively similar than gemma2
  • 1 reply
·
julien-c 
posted an update 1 day ago
view post
Post
1754
Important notice 🚨

For Inference Providers who have built support for our Billing API (currently: Fal, Novita, HF-Inference – with more coming soon), we've started enabling Pay as you go (=PAYG)

What this means is that you can use those Inference Providers beyond the free included credits, and they're charged to your HF account.

You can see it on this view: any provider that does not have a "Billing disabled" badge, is PAYG-compatible.
m-ric 
posted an update 3 days ago
view post
Post
755
Our new Agentic leaderboard is now live!💥

If you ever asked which LLM is best for powering agents, we've just made a leaderboard that ranks them all! Built with @albertvillanova , this ranks LLMs powering a smolagents CodeAgent on subsets of various benchmarks. ✅

🏆 GPT-4.5 comes on top, even beating reasoning models like DeepSeek-R1 or o1. And Claude-3.7-Sonnet is a close second!

The leaderboard also allows you to show the scores of vanilla LLMs (without any agentic setup) on the same benchmarks: this shows the huge improvements brought by agentic setups. 💪

(Note that results will be added manually, so the leaderboard might not always have the latest LLMs)
  • 1 reply
·
clem 
posted an update 5 days ago
view post
Post
6955
I was chatting with @peakji , one of the cofounders of Manu AI, who told me he was on Hugging Face (very cool!).

He shared an interesting insight which is that agentic capabilities might be more of an alignment problem rather than a foundational capability issue. Similar to the difference between GPT-3 and InstructGPT, some open-source foundation models are simply trained to 'answer everything in one response regardless of the complexity of the question' - after all, that's the user preference in chatbot use cases. Just a bit of post-training on agentic trajectories can make an immediate and dramatic difference.

As a thank you to the community, he shared 100 invite code first-come first serve, just use “HUGGINGFACE” to get access!
·
clem 
posted an update 6 days ago
clem 
posted an update 9 days ago
view post
Post
5866
Super happy to welcome Nvidia as our latest enterprise hub customer. They have almost 2,000 team members using Hugging Face, and close to 20,000 followers of their org. Can't wait to see what they'll open-source for all of us in the coming months!

Nvidia's org: https://huggingface.co/nvidia
Enterprise hub: https://huggingface.co/enterprise
TimDarcet 
authored a paper 10 days ago
samchain 
posted an update 16 days ago
view post
Post
1824
NLP for economics 1.1 is out !

Following the 1.0 collection, I release the 1.1 version with an updated dataset for sentence similarity as well as a raw dataset from central bankers speeches.

The newest model is econo-sentence-v2 is a new version of a sentence-transformers model based on EconoBert ! It gets better results with a nuance on similarity.

If you're an economist looking for useful tools, don't hesitate to check it out !
m-ric 
posted an update 17 days ago
view post
Post
4726
We now have a Deep Research for academia: SurveyX automatically writes academic surveys nearly indistinguishable from human-written ones 🔥

Researchers from Beijing and Shanghai just published the first application of a deep research system to academia: their algorithm, given a question, can give you a survey of all papers on the subject.

To make a research survey, you generally follow two steps, preparation (collect and organize papers) and writing (outline creation, writing, polishing). Researchers followed the same two steps and automated them.

🎯 For the preparation part, a key part is find all the important references on the given subject.
Researchers first cast a wide net of all relevant papers. But then finding the really important ones is like distilling knowledge from a haystack of information. To solve this challenge, they built an “AttributeTree” object that structures key information from citations. Ablating these AttributeTrees significantly decreased structure and synthesis scores, so they were really useful!

📝 For the writing part, key was to get a synthesis that's both short and true. This is not easy to get with LLMs! So they used methods like LLM-based deduplication to shorten the too verbose listings made by LLMs, and RAG to grab original quotes instead of made-up ones.

As a result, their system outperforms previous approaches by far!

As assessed by LLM-judges, the quality score os SurveyX even approaches this of human experts, with 4.59/5 vs 4.75/5 🏆

I advise you to read the paper, it's a great overview of the kind of assistants that we'll get in the short future! 👉 SurveyX: Academic Survey Automation via Large Language Models (2502.14776)
Their website shows examples of generated surveys 👉 http://www.surveyx.cn/
sayandsarkar 
authored a paper 17 days ago
clem 
posted an update 23 days ago
view post
Post
2819
What are the best organizations to follow on @huggingface ?

On top of my head:
- Deepseek (35,000 followers): https://huggingface.co/deepseek-ai
- Meta Llama (27,000 followers): https://huggingface.co/meta-llama
- Black Forrest Labs (11,000 followers): https://huggingface.co/black-forest-labs
- OpenAI (5,000 followers): https://huggingface.co/openai
- Nvidia (16,000 followers): https://huggingface.co/nvidia
- MIcrosoft (9,000 followers): https://huggingface.co/microsoft
- AllenAI (2,000 followers): https://huggingface.co/allenai
- Mistral (5,000 followers): https://huggingface.co/mistralai
- XAI (600 followers): https://huggingface.co/xai-org
- Stability AI (16,000 followers): https://huggingface.co/stabilityai
- Qwen (16,000 followers): https://huggingface.co/Qwen
- GoogleAI (8,000 followers): https://huggingface.co/google
- Unsloth (3,000 followers): https://huggingface.co/unsloth
- Bria AI (4,000 followers): https://huggingface.co/briaai
- NousResearch (1,300 followers): https://huggingface.co/NousResearch

Bonus, the agent course org with 17,000 followers: https://huggingface.co/agents-course
  • 1 reply
·
m-ric 
posted an update 23 days ago
view post
Post
3025
Less is More for Reasoning (LIMO): a 32B model fine-tuned with 817 examples can beat o1-preview on math reasoning! 🤯

Do we really need o1's huge RL procedure to see reasoning emerge? It seems not.
Researchers from Shanghai Jiaotong University just demonstrated that carefully selected examples can boost math performance in large language models using SFT —no huge datasets or RL procedures needed.

Their procedure allows Qwen2.5-32B-Instruct to jump from 6.5% to 57% on AIME and from 59% to 95% on MATH, while using only 1% of the data in previous approaches.

⚡ The Less-is-More Reasoning Hypothesis:
‣ Minimal but precise examples that showcase optimal reasoning patterns matter more than sheer quantity
‣ Pre-training knowledge plus sufficient computational resources at inference levels up math skills

➡️ Core techniques:
‣ High-quality reasoning chains with self-verification steps
‣ 817 handpicked problems that encourage deeper reasoning
‣ Enough inference-time computation to allow extended reasoning

💪 Efficiency gains:
‣ Only 817 examples instead of 100k+
‣ 40.5% absolute improvement across 10 diverse benchmarks, outperforming models trained on 100x more data

This really challenges the notion that SFT leads to memorization rather than generalization! And opens up reasoning to GPU-poor researchers 🚀

Read the full paper here 👉  LIMO: Less is More for Reasoning (2502.03387)
sayakpaul 
posted an update 23 days ago
view post
Post
3080
Inference-time scaling meets Flux.1-Dev (and others) 🔥

Presenting a simple re-implementation of "Inference-time scaling diffusion models beyond denoising steps" by Ma et al.

I did the simplest random search strategy, but results can potentially be improved with better-guided search methods.

Supports Gemini 2 Flash & Qwen2.5 as verifiers for "LLMGrading" 🤗

The steps are simple:

For each round:

1> Starting by sampling 2 starting noises with different seeds.
2> Score the generations w.r.t a metric.
3> Obtain the best generation from the current round.

If you have more compute budget, go to the next search round. Scale the noise pool (2 ** search_round) and repeat 1 - 3.

This constitutes the random search method as done in the paper by Google DeepMind.

Code, more results, and a bunch of other stuff are in the repository. Check it out here: https://github.com/sayakpaul/tt-scale-flux/ 🤗
clem 
posted an update 24 days ago
view post
Post
3483
We crossed 1B+ tokens routed to inference providers partners on HF, that we released just a few days ago.

Just getting started of course but early users seem to like it & always happy to be able to partner with cool startups in the ecosystem.

Have you been using any integration and how can we make it better?

https://huggingface.co/blog/inference-providers
m-ric 
posted an update 27 days ago
view post
Post
2903
𝗚𝗿𝗲𝗮𝘁 𝗳𝗲𝗮𝘁𝘂𝗿𝗲 𝗮𝗹𝗲𝗿𝘁: you can now share agents to the Hub! 🥳🥳

And any agent pushed to Hub get a cool Space interface to directly chat with it.

This was a real technical challenge: for instance, serializing tools to export them meant that you needed to get all the source code for a tool, verify that it was standalone (not relying on external variables), and gathering all the packages required to make it run.

Go try it out! 👉 https://github.com/huggingface/smolagents
  • 2 replies
·
m-ric 
posted an update 27 days ago
view post
Post
2514
For those who haven't come across it yet, here's a handy trick to discuss an entire GitHub repo with an LLM:

=> Just replace "github" with "gitingest" in the url, and you get the whole repo as a single string that you can then paste in your LLMs
m-ric 
posted an update 29 days ago
view post
Post
4821
"𝟮𝟬𝟮𝟱 𝘄𝗶𝗹𝗹 𝗯𝗲 𝘁𝗵𝗲 𝘆𝗲𝗮𝗿 𝗼𝗳 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀": this statement has often been made, here are numbers to support it.

I've plotted the progress of AI agents on GAIA test set, and it seems they're headed to catch up with the human baseline in early 2026.

And that progress is still driven mostly by the improvement of base LLMs: progress would be even faster with fine-tuned agentic models.
m-ric 
posted an update about 1 month ago
view post
Post
3735
𝗔𝗱𝘆𝗲𝗻'𝘀 𝗻𝗲𝘄 𝗗𝗮𝘁𝗮 𝗔𝗴𝗲𝗻𝘁𝘀 𝗕𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 𝘀𝗵𝗼𝘄𝘀 𝘁𝗵𝗮𝘁 𝗗𝗲𝗲𝗽𝗦𝗲𝗲𝗸-𝗥𝟭 𝘀𝘁𝗿𝘂𝗴𝗴𝗹𝗲𝘀 𝗼𝗻 𝗱𝗮𝘁𝗮 𝘀𝗰𝗶𝗲𝗻𝗰𝗲 𝘁𝗮𝘀𝗸𝘀! ❌

➡️ How well do reasoning models perform on agentic tasks? Until now, all indicators seemed to show that they worked really well. On our recent reproduction of Deep Search, OpenAI's o1 was by far the best model to power an agentic system.

So when our partner Adyen built a huge benchmark of 450 data science tasks, and built data agents with smolagents to test different models, I expected reasoning models like o1 or DeepSeek-R1 to destroy the tasks at hand.

👎 But they really missed the mark. DeepSeek-R1 only got 1 or 2 out of 10 questions correct. Similarly, o1 was only at ~13% correct answers.

🧐 These results really surprised us. We thoroughly checked them, we even thought our APIs for DeepSeek were broken and colleagues Leandro Anton helped me start custom instances of R1 on our own H100s to make sure it worked well.
But there seemed to be no mistake. Reasoning LLMs actually did not seem that smart. Often, these models made basic mistakes, like forgetting the content of a folder that they had just explored, misspelling file names, or hallucinating data. Even though they do great at exploring webpages through several steps, the same level of multi-step planning seemed much harder to achieve when reasoning over files and data.

It seems like there's still lots of work to do in the Agents x Data space. Congrats to Adyen for this great benchmark, looking forward to see people proposing better agents! 🚀

Read more in the blog post 👉 https://huggingface.co/blog/dabstep
eliebak 
authored 2 papers about 1 month ago