


Building @zml_ai (and we're hiring), ex @zenly, ex Exalead, ex @google. Skydiver and wingsuiter.


7A1B is actually very interesting, congratulations on the launch
OVH launches Olala, its first family of language models. First of the family, Olala mini 7A1B is a compact model with 7 billion parameters, 1 billion of them activated, trained from scratch on European public supercomputers. It relies on an innovative architecture, runs even without a GPU, and will be released as open source. A model has become a strategic building block. We chose to build it rather than depend on it. This model proves that European companies can innovate and train agentic LLMs from scratch. Next step: Olala v1.1 by the end of the year with improved performance, then Olala v2.0, which will exceed 100 billion parameters. More info ➡️ lnkd.in/gyreQ8nz Come and discover Olala mini 7A1B and the details of our cloud and AI innovations at the OVHcloud AI Summit, on 19 November. #OVHcloudAISummit

Backprop Summit you're a bunch of bozos. For the first time in my career, I couldn't get to an event because the venue was at capacity even though we had registered. So were a lot of other folks who also had registered. I was actually invited to the event and drove from Santa Clara to SF to attend.

i don't think i've seen a open weight release waited that much by the community
📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery! Qwen4-Max is built on a new MoE architecture with native vision and a 1M-token context window. Highlights: - Outperforms Qwen3.8-Max with up to ~50% fewer tokens, while average cost per task drops over 60% in our internal evaluation. - Advances real-world productivity through greater proficiency with professional tools, stronger long-horizon execution, and higher-quality deliverables. - Handles real tasks, including 3D modeling, game development, manufacturing, finance, law, healthcare, and frontier scientific discovery. - $1.5/M tokens for input, $4.5/M tokens for output, and $0.075/M tokens for implicit caching — all lower than Qwen3.8-Max, with implicit caching down by 70%. The API is live on QwenCloud. You can also try Qwen4-Max on Qwen Studio, QwenWork and Qoder. Start building today! 🚀 - Blog: qwen.ai/blog?id=qwen4-max - API: qwencloud.com/models/qwen4-m… - Token Plan: qwencloud.com/pricing/token-… More to come! The open weights of smaller Qwen4 models will be released. Stay tuned. 👀

see you all tomorrow at InferCon in San Francisco

gee i wonder how this is going to unfold </s>
I noticed that the Crimea is marked as russian territory on the Omarchy Meetup website. So, asked grok to fix it with the links to the White House official documents. I hope the PR will be merged! github.com/omacom/omarchy-si…

help me find a reason to integrate libtorrent in zml

"semantic" is the word i probably use the most in my job

And here is @zml_ai with CUDA running natively on Windows with a RTX 4090 via PCIe passthrough. (recording over VNC)

And here is @zml_ai with CUDA running natively on Windows with a RTX 4090 via PCIe passthrough. (recording over VNC)
and here is @zml_ai running Llama 3.2 1B *fully natively on @Windows* (CPU for now, it's a demo). little nugget: it was entirely cross compiled from macOS using @hermeticbuild's LLVM

our greatest release ever!
Announcing the release of zml/llmd 20261112.0. This week, we are bringing DSpine, more speculative decoding options, faster DeepSeek prefill, Apple M1 and M2 support, more control over sampling. And of course, lots of bugfixes! zml.ai/posts/llmd-20261112.0…

Can someone get us access to Ascend 950 ?

released today, blog post tomorrow
and here is DSpine running in zml/llmd, seeing upwards of 450+ tok/s on prompts like "write a website generator in python", holy hell

challenge accepted
Lightpanda agent + StartLux decision local model vs Claude Haiku Look forward to testing with llmd @zml_ai 🫡

seems i need to repost this
the Jensen scandal wouldn't work in France

hi, as a French person: what in the fuck ?
Jensen Huang must go We argue in @TheEconomist that his position is untenable "Like many powerful leaders, he came to believe that he was too important for the normal rules to apply to him." But he is not the only one. Howard Lutnick has failed to stem the flow of chips to China. More heads should roll