Hugging Face's logo

smol-explorers
community
Activity Feed

AI & ML interests

None defined yet.

Recent Activity

manu  authored a paper about 2 months ago
Should We Still Pretrain Encoders with Masked Language Modeling?
thomwolf  authored a paper about 2 months ago
FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language
loubnabnl  authored a paper 3 months ago
The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text
View all activity

eliebak 
posted an update 1 day ago
view post
Post
260
Motif 2.6B tech report is pretty insane, first time i see a model with differential attention and polynorm trained at scale!

> It's trained on 2.5T of token, with a "data mixture schedule" to continuously adjust the mixture over training.
> They use WSD with a "Simple moving average" averaging the last 6 ckpt every 8B token.
> They trained on Finemath, Fineweb2, DCLM, TxT360.
> Lot of details in the finetuning data they used, for instance they used EvolKit and did some "dataset fusion" to have more compressed knowledge into the data.
> They mention they also tried Normalized GPT, QK-Norm and Cross Layer Attention.

Motif-Technologies/Motif-2.6B
hlarcher 
posted an update 25 days ago
view post
Post
231
GH200 cooking time 🧑‍🍳🔥!

We just updated GPU-fryer 🍳 to run on Grace Hopper Superchip (GH200) - fully optimized for ARM-based systems!
With this release, we switched to cuBLASLt to support running FP8 benchmarks. You can monitor GPU throttling, TFLOPS outliers, HBM memory health, and ensure that you get the most of your hardware setup.
Perfect for stress testing and tuning datacenter GPUs.

Check it out on Github 👉 https://github.com/huggingface/gpu-fryer
andito 
posted an update about 1 month ago
view post
Post
2782
Many VLMs claim to process hours of video. But can they follow the story?🤔
Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!⏳

We test three skills that matter for real-world use:
🔎 Localized Retrieval: Find a specific action.
🧩 Information Synthesis: Piece together scattered clues.
🏃 Fine-Grained Perception: Analyze detailed motion (e.g., count how many times a person swings an axe).

The results are in, and they're revealing. Only Gemini 2.5 pro handles 1-hour-long videos.
Performance drops sharply with duration, proving that long video understanding is still challenging. We've found the breaking points—now the community can start fixing them.📈

Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI.

📖 Blog:
https://huggingface.co/blog/timescope-video-lmm-benchmark
👩‍💻 Leaderboard & Demo: Apollo-LMMs/TimeScope
📊 Dataset: Apollo-LMMs/TimeScope
⚙️ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
eliebak 
posted an update about 1 month ago
view post
Post
4668
Kimi K2 tech report is full of gems as always. Here are my notes on it:

> MuonClip: Pretty crazy how after 70k the training stabilizes and the QK-clip is basically inactive. There is also no loss in perf with QK-clip which is not trivial at all (at small scale but with aggressive threshold). Also a cool explanation of why muon makes the logit explode in appendix E (tl;dr is that muon makes the singular value of the update matrix higher)
> Sparsity scaling laws to justify their ratio, they have a very solid training infra that allows the model to be trained at this sparsity level, they could have increased even more but as sparsity increases the training becomes less efficient.
> They diminish the number of attention heads to make it more efficient for long context since attention heads are a big bottleneck for long context. They also remove 2 of the 3 "first dense" layers in the dsv3 arch.

With the sparsity and attention heads (divided by 2) they achieve 83% increased flops compared to deepseek v3 arch at 128k.

> Data: Rephrasing is KEY. They do a lot more synthetic data generation and rephrase their corpus to have different styles, for longer documents they do it by chunk. I'm (half) surprised by the fact that ONLY 1 epoch (assuming same number of training tokens I think?) of data rephrased 10 times has better accuracy than 10 epochs of the same data rephrased once.
> They do rewriting for Math and Knowledge, for Math they apply the ShallowMath recipe and instruct the model to rephrase in a "learning note" style
> They talk about diversity and probably have some internal stuff/eval to test that, as always still a bit unclear for me how to properly measure that.

The infra is also very nice, quick summary:
> PP=16 (1F1B schedule, a bit custom), EP=16, zero1
> No FP8 computation but for storage of specific layers, selective recomputation for inexpensive block, activation offloading to CPU
manu 
authored a paper about 2 months ago
andito 
posted an update about 2 months ago
view post
Post
3988
🧠👁️ Can AI visualize solutions?

Humans often solve visual problems by sketching ideas in our minds. What if Vision-Language Models (VLMs) could do something similar, not by generating full images, but by using internal “mental sketches”?

That’s the idea behind Mirage, a new framework that empowers VLMs to reason using latent visual tokens. Instead of just thinking in words, Mirage mixes in abstract visual representations that help the model solve complex tasks.

These aren't photorealistic images. They're compact, internal representations optimized purely to support reasoning.

🔧 Mirage is trained in two phases:

1) Grounding: It learns to produce latent tokens anchored in real images.
2) Refinement: The model drops the images and learns to generate visual tokens on its own.

📈 And yes, it works!
On challenging benchmarks like Visual Spatial Planning, Jigsaw puzzles, and Spatial Attention Tasks, Mirage clearly outperforms GPT-4o and other strong baselines.
Smart sketches > empty words.

By mimicking the way humans visualize solutions, Mirage gives AI a new kind of imagination, one that’s faster, more efficient, and more human-like.
Kudos to the teams at UMass Amherst and MIT behind this exciting work.
Check the paper: Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens (2506.17218)
·
thomwolf 
authored a paper about 2 months ago
loubnabnl 
authored a paper 3 months ago
eliebak 
authored a paper 3 months ago
tonywu71 
authored a paper 3 months ago
danaaubakirova 
posted an update 3 months ago
danaaubakirova 
authored a paper 3 months ago
thomwolf 
authored a paper 3 months ago
andito 
authored a paper 3 months ago
danaaubakirova 
authored a paper 3 months ago
manu 
authored a paper 3 months ago
loubnabnl 
posted an update 3 months ago
thomwolf 
posted an update 4 months ago
view post
Post
6410
If you've followed the progress of robotics in the past 18 months, you've likely noticed how robotics is increasingly becoming the next frontier that AI will unlock.

At Hugging Face—in robotics and across all AI fields—we believe in a future where AI and robots are open-source, transparent, and affordable; community-built and safe; hackable and fun. We've had so much mutual understanding and passion working with the Pollen Robotics team over the past year that we decided to join forces!

You can already find our open-source humanoid robot platform Reachy 2 on the Pollen website and the Pollen community and people here on the hub at pollen-robotics

We're so excited to build and share more open-source robots with the world in the coming months!
  • 1 reply
·
thomwolf 
authored a paper 5 months ago
hlarcher 
authored a paper 5 months ago