Загрузка видео...

Не удалось загрузить видео

На главную

Meta is back with Muse Glimmer: a 30B open-source multimodal model built for local, agentic use. HF is shipping day-0 support and I built a few demos to see what it can do. First: we gave Glimmer tools and asked it to quantize itself.

51,554 просмотров • 2 месяцев назад •via X (Twitter)

Комментарии: 11

Фото профиля Ben Burtenshaw
Ben Burtenshaw2 месяцев назад

Next: “Muse Glimmer, deploy yourself.” It deploys to a protected Hugging Face Inference Endpoint, verifies health and chat completion, then connects its claw to the endpoint. You can also just do this with:

Фото профиля Ben Burtenshaw
Ben Burtenshaw2 месяцев назад

If you just want to try the model, I set up this research agent on the hub:

Фото профиля Ben Burtenshaw
Ben Burtenshaw2 месяцев назад

Finally, I got it to optimize its own deployment for H100 architecture.

Фото профиля Noé Flandre
Noé Flandre2 месяцев назад

Sounds like a solid release. Very exciting to see Meta coming back in the OSS game. Let’s see how it holds against 3.8

Фото профиля Ben Burtenshaw
Ben Burtenshaw2 месяцев назад

Yes! So excited to see what people build with it.

Фото профиля Fajar M Reza
Fajar M Reza2 месяцев назад

Open-weight multimodal models matter when local agents can inspect tools and images.

Фото профиля Shantanu Goel
Shantanu Goel2 месяцев назад

is this in hermes?

Фото профиля Ben Burtenshaw
Ben Burtenshaw2 месяцев назад

openclaw, but I've done similar stuff in Hermes and it works.

Фото профиля Marcus Supersonic
Marcus Supersonic2 месяцев назад

Self-quantization is a neat demo. Whether agentic tool use holds up on consumer hardware is the actual question.

Фото профиля Ryan
Ryan2 месяцев назад

@grok what pc specs do you recommend to run this smoothly all day.

Фото профиля Adel Bucetta
Adel Bucetta2 месяцев назад

haven't seen anyone try to give an AI the task of optimizing its own architecture yet. that's a whole new level of self-awareness

Похожие видео

First impressions on Muse Glimmer! It's incredibly fast for a dense model, currently running an average of 208tps with a max of 274tps on a single 5090 with their DFLASH config. Comparatively, though, both using Open Code, Qwopus Coder (with thinking off) produced a much better shark survival game than the one I got from Glimmer. Meta's new dense model is currently just lacking some HTML canvas taste, but this is something that can be added via SFT as long as the model is stable and capable from a back-end programming perspective. And it seems to be, without a doubt. The big kicker here is that I ran this at extra high thinking, and it did not take long at all to run. Our current local leader, Qwen 27B 3.6, has a tendency to overthink, but with glimmer, that is not the case. Right now, my recommendation for general local programming (Apps, Games, Websites, Visual Tools) in this class is still Qwopus Coder with thinking disabled, or Qwopus Fusion with thinking enabled. Of course Shark Survival is a very basic domain-specific test, but I find that the result scales very well across many domains. If we're going to be shipping apps generated entirely locally, visual taste is somewhat of a bare minimum requirement, solely in my opinion, and Qwen's models in this class offer significantly more at the moment. That's actually why I initially started getting into finetuning with Qwen 3.5, they were the first base that was able to do really good front-end with some opus-trace fine-tuning. Qwen 3.6 has taste even in the base model, and we know Qwen 3.8 is going to blow us all away! Regardless, this looks like a very tempting new base model. As a first offering from Meta in this class for a long time, I am incredibly impressed and elated to have it. We now finally have a proper Single GPU frontier race, instead of us just begging Qwen for more releases. Single GPU open frontier model race is a VERY good thing. Please keep pushing Meta

Kyle Hessling

17,485 просмотров • 2 месяцев назад