9 comments

  • brainless 40 minutes ago ago

    I am sorry I did not understand all of it. But, would this allow running large MoE LLMs on a local network with experts spread out over multiple cheaper GPUs (or even CPUs)? This would perhaps be more useful than over the Internet, within offices for example.

  • s2l 39 minutes ago ago

    What if one wishes to use various busybox nodes within the house?

    All the iot devices contributing to matmul but within a LAN?

  • snovv_crash an hour ago ago

    Cool idea. How do you handle temperature in the verification?

    • vforno an hour ago ago

      Ho thanks for the comment. Verification does not depend on temperature. Expert execution is deterministic (pure matmul). LUMABRI_VERIFY=N re-runs N% of the calls on a second replica and requires byte-identical output. Temperature (and sampling) happens only on the chatter, after the experts return their activations. So it can be any value (0, 0.7, 1.2…) without affecting the verification contract.

      • stymaar 3 minutes ago ago

        > Expert execution is deterministic (pure matmul).

        Isn't that only true in theory but wrong in practice due to floating points?

  • Scaled an hour ago ago

    Wasn't a p2p ai model how skynet got started?

    J/k, this looks cool :)

  • ptsneves 2 hours ago ago

    P2P inference sounds really nice and it would bring us back to the 2000s culture if not for the fact inference is already so cheap. Even so it is an insurance policy if the cloud providers or governments have ideas of further censoring and monitoring chats.

  • RIshabh235 3 hours ago ago

    great work

    • vforno 2 hours ago ago

      Thanks for the support!