10 comments

  • Havoc 2 minutes ago
    Sounds interesting.

    Got a bunch of errors trying to run it on CPU though. Very likely connected to me running this in a container (unpriv LXC), but figured for 26M CPU would suffice.

    https://pastebin.com/PYZJKTNk

  • simonw 51 minutes ago
    Suggestion: publish a live demo of the "needle playground". It's small enough that it should be pretty cheap to run this on a little VPS somewhere!
    • quantumleaper 38 minutes ago
      Should be quick and easy with WebGPU, too.
      • simonw 14 minutes ago
        That's an even better idea, I bet this could run in Transformers.js.
      • ilaksh 27 minutes ago
        Good idea. Could you make that.
    • HenryNdubuaku 45 minutes ago
      thanks, yeah, the problem is just handling scale, we don't have the infra ready to go, but anyone can do that. Its easy for people to run on their laptops straight up. Will try the VPS route.
  • kristopolous 17 minutes ago
    That M versus B is way too subtle. 0.026B is my suggestion
  • simonw 1 hour ago
    Looks like you need to open up access to https://huggingface.co/Cactus-Compute/datasets/needle-tokeni... - I get this error when trying to run the steps in your README:

    > Repository Not Found for url: http s://huggingface.co/api/datasets/Cactus-Compute/needle-tokenizer/revision/main.

  • ilaksh 1 hour ago
    Hmm.. this might make it feasible to build something like a command line program where you can optionally just specify the arguments in natural language. Although I know people will object to including an extra 14 MB and the computation for "parsing" and it could be pretty bad if everyone started doing that.

    But it's really interesting to me that that may be possible now. You can include a fine-tuned model that understands how to use your program.

    E.g. `> toolcli what can you do` runs `toolcli --help summary`, `toolcli add tom to teamfutz group` = `toolcli --gadd teamfutz tom`

    • HenryNdubuaku 1 hour ago
      So Needle is trained for INT4, what you see in the playground is INT4, only 14MB, same challenge though.
      • ilaksh 1 hour ago
        Oh gotcha. Fixed my comment.
  • murkt 35 minutes ago
    Can this be a Siri-like core? Set me a timer, tell me what’s the weather, etc. Here is transcribed text and available list of tools for the model to call, and voice the output.
  • cmrdporcupine 53 minutes ago
    This is very cool I'm going to try to carve out some time to try building this into my MOO system ( https://codeberg.org/timbran/moor / https://timbran.org/moor.html ) as alternative command parser front end.
    • Balinares 6 minutes ago
      Man, I love that there are still people writing new MOO servers in 2026. Any game out there already running on mooR?
    • HenryNdubuaku 44 minutes ago
      Thanks, let us know how it goes!
  • nhattruongadm 55 minutes ago
    [flagged]
  • abhijithbabu 2 hours ago
    [flagged]
  • ac29 36 minutes ago
    FYI, distilling Gemini is explicitly against the ToS:

    "You may not use the Services to develop models that compete with the Services (e.g., Gemini API or Google AI Studio). You also may not attempt to reverse engineer, extract or replicate any component of the Services, including the underlying data or models (e.g., parameter weights)."

    • Havoc 5 minutes ago
      Yeah I think Google should shove that somewhere. They effectively distilled all the internet's knowledge into these models...without asking & without permission
    • ilaksh 29 minutes ago
      I think GLM 5.1 or Kimi 2.6 could substitute for this type of purpose.
    • ForHackernews 17 minutes ago
      So is copying all the books in the world.
    • vablings 34 minutes ago
      Oh no! They stole the model weights! Distillation "attacks" is such bullshit
    • xgulfie 10 minutes ago
      This is being downvoted but it's worth noting if only for the "be careful" aspect.

      That said, we need more people distilling models IMO, just be ready for a C&D and a ban