Hugging Face's logo

Hugging Face H4

Enterprise
company
https://github.com/huggingface/alignment-handbook
Activity Feed

AI & ML interests

Aligning LLMs to be helpful, honest, harmless, and huggy (H4)

Recent Activity

lewtun  new activity 6 days ago
HuggingFaceH4/blogpost-scaling-test-time-compute:Use HF papers
lewtun  new activity 6 days ago
HuggingFaceH4/blogpost-scaling-test-time-compute:Update app/src/index.html
qgallouedec  new activity 6 days ago
HuggingFaceH4/blogpost-scaling-test-time-compute:Use HF papers
View all activity

HuggingFaceH4's activity

regisss 
posted an update 4 days ago
freddyaboulton 
posted an update 4 days ago
merve 
posted an update 4 days ago
view post
Post
1992
Aya by Cohere For AI can now see! 👀

C4AI community has built Maya 8B, a new open-source multilingual VLM built on SigLIP and Aya 8B 🌱 works on 8 languages! 🗣️

The authors extend Llava dataset using Aya's translation capabilities with 558k examples!
ry it here kkr5155/maya_demo

Dataset maya-multimodal/pretrain

Model maya-multimodal/maya 👏
kudos @nahidalam and team
  • 1 reply
·
freddyaboulton 
posted an update 5 days ago
merve 
posted an update 5 days ago
view post
Post
2589
Apollo is a new family of open-source video language models by Meta, where 3B model outperforms most 7B models and 7B outperforms most 30B models 🧶

✨ the models come in 1.5B https://huggingface.co/Apollo-LMMs/Apollo-1_5B-t32, 3B https://huggingface.co/Apollo-LMMs/Apollo-3B-t32 and 7B https://huggingface.co/Apollo-LMMs/Apollo-7B-t32 with A2.0 license, based on Qwen1.5 & Qwen2
✨ the authors also release a benchmark dataset https://huggingface.co/spaces/Apollo-LMMs/ApolloBench

The paper has a lot of experiments (they trained 84 models!) about what makes the video LMs work ⏯️

Try the demo for best setup here https://huggingface.co/spaces/Apollo-LMMs/Apollo-3B
they evaluate sampling strategies, scaling laws for models and datasets, video representation and more!
> The authors find out that whatever design decision was applied to small models also scale properly when the model and dataset are scaled 📈 scaling dataset has diminishing returns for smaller models
> They evaluate frame sampling strategies, and find that FPS sampling is better than uniform sampling, and they find 8-32 tokens per frame optimal
> They also compare image encoders, they try a variation of models from shape optimized SigLIP to DINOv2
they find google/siglip-so400m-patch14-384 to be most powerful 🔥
> they also compare freezing different parts of models, training all stages with some frozen parts give the best yield

They eventually release three models, where Apollo-3B outperforms most 7B models and Apollo 7B outperforms 30B models 🔥
  • 3 replies
·
lewtun 
posted an update 6 days ago
view post
Post
6328
We outperform Llama 70B with Llama 3B on hard math by scaling test-time compute 🔥

How? By combining step-wise reward models with tree search algorithms :)

We show that smol models can match or exceed the performance of their much larger siblings when given enough "time to think"

We're open sourcing the full recipe and sharing a detailed blog post.

In our blog post we cover:

📈 Compute-optimal scaling: How we implemented DeepMind's recipe to boost the mathematical capabilities of open models at test-time.

🎄 Diverse Verifier Tree Search (DVTS): An unpublished extension we developed to the verifier-guided tree search technique. This simple yet effective method improves diversity and delivers better performance, particularly at large test-time compute budgets.

🧭 Search and Learn: A lightweight toolkit for implementing search strategies with LLMs and built for speed with vLLM

Here's the links:

- Blog post: HuggingFaceH4/blogpost-scaling-test-time-compute

- Code: https://github.com/huggingface/search-and-learn

Enjoy!
  • 2 replies
·
lewtun 
in HuggingFaceH4/blogpost-scaling-test-time-compute 6 days ago

Use HF papers

1
#2 opened 6 days ago by
qgallouedec

Update app/src/index.html

1
#1 opened 6 days ago by
qgallouedec
qgallouedec 
in HuggingFaceH4/blogpost-scaling-test-time-compute 6 days ago

Use HF papers

1
#2 opened 6 days ago by
qgallouedec
lewtun 
updated a Space 6 days ago
qgallouedec 
updated a Space 6 days ago
qgallouedec 
in HuggingFaceH4/blogpost-scaling-test-time-compute 6 days ago

Update app/src/index.html

1
#1 opened 6 days ago by
qgallouedec
lewtun 
updated a dataset 6 days ago
lewtun 
updated a collection 9 days ago
freddyaboulton 
posted an update 10 days ago
view post
Post
1786
Version 0.0.21 of gradio-pdf now properly loads chinese characters!
yjernite 
posted an update 10 days ago
view post
Post
2019
🇪🇺 Policy Thoughts in the EU AI Act Implementation 🇪🇺

There is a lot to like in the first draft of the EU GPAI Code of Practice, especially as regards transparency requirements. The Systemic Risks part, on the other hand, is concerning for both smaller developers and for external stakeholders.

I wrote more on this topic ahead of the next draft. TLDR: more attention to immediate large-scale risks and to collaborative solutions supported by evidence can help everyone - as long as developers disclose sufficient information about their design choices and deployment contexts.

Full blog here, based on our submitted response with @frimelle and @brunatrevelin :

https://huggingface.co/blog/yjernite/eu-draft-cop-risks#on-the-proposed-taxonomy-of-systemic-risks
  • 2 replies
·
freddyaboulton 
posted an update 10 days ago
view post
Post
1501
Hello Llama 3.2! 🗣️🦙

Build a Siri-like coding assistant that responds to "Hello Llama" in 100 lines of python! All with Gradio, webRTC 😎

freddyaboulton/hey-llama-code-editor
Narsil 
posted an update 10 days ago
view post
Post
904
Performance leap: TGI v3 is out. Processes 3x more tokens, 13x faster than vLLM on long prompts. Zero config !



3x more tokens.

By reducing our memory footprint, we’re able to ingest many more tokens and more dynamically than before. A single L4 (24GB) can handle 30k tokens on llama 3.1-8B, while vLLM gets barely 10k. A lot of work went into reducing the footprint of the runtime and its effect are best seen on smaller constrained environments.
13x faster

On long prompts (200k+ tokens) conversation replies take 27.5s in vLLM, while it takes only 2s in TGI. How so ? We keep the initial conversation around, so when a new reply comes in, we can answer almost instantly. The overhead of the lookup is ~5us. Thanks @Dani ël de Kok for the beast data structure.
Zero config

That’s it. Remove all the flags your are using and you’re likely to get the best performance. By evaluating the hardware and model, TGI carefully selects automatic values to give best performance. In production, we don’t have any flags anymore in our deployments. We kept all existing flags around, they may come in handy in niche scenarios.

Read more: https://huggingface.co/docs/text-generation-inference/conceptual/chunking