19 comments

  • simonw 2 hours ago
    If you want to try out out the GGUFs from https://huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gguf#th... be aware that you need Prism's llama.cpp fork to get them to work, from https://github.com/PrismML-Eng/llama.cpp/releases/tag/prism-...

    This should work:

      cd /tmp
    
      # Get the Prism macOS runtime
      curl -fL https://github.com/PrismML-Eng/llama.cpp/releases/download/prism-b10685-7dffb15/llama-prism-b10685-7dffb15-bin-macos-arm64.tar.gz -o bonsai-runtime.tar.gz
      tar -xzf bonsai-runtime.tar.gz
    
      # Get the ~5.95 GB GGUF model:
      curl -fL https://huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gguf/resolve/main/Ternary-Bonsai-2-27B-PTQ1_0.gguf -o Ternary-Bonsai-2-27B-PTQ1_0.gguf
    
      # Run the server, I used port 8331
      ./llama-prism-b10685-7dffb15/llama-server \
        -m Ternary-Bonsai-2-27B-PTQ1_0.gguf \
        --port 8331 -ngl 99 -fa on -c 32768
    
    Then open http://localhost:8331 for the (very good) baked in llama-server web UI... or run a prompt via the API like this:

      uvx llm openai endpoint http://127.0.0.1:8331/v1 \
        --model bonsai-2-27b --responses hi
    
    That's running at ~20 token/second for me on an M5 Pro (after a server restart I got 44 token/second, not sure why), but I'm pretty sure something isn't working right, on startup the server said "ggml_metal_device_init: - the tensor API is not supported in this environment - disabling".
    • rahimnathwani 36 minutes ago
      If you want to download the gguf to your regular huggingface cache directory instead of to /tmp, you can download the model and run the server in one step:

        export HF_TOKEN=xxx # optional, speeds up the download
        
        ./llama-prism-b10685-7dffb15/llama serve \
          -hf prism-ml/Ternary-Bonsai-2-27B-gguf:PTQ1_0 \
          --port 8331 -ngl 99 -fa on -c 32768
    • simonw 1 hour ago
      I used that to Generate an SVG of a pelican riding a bicycle:

      https://tools.simonwillison.net/markdown-svg-renderer?url=ht...

      It took 18 minutes 20 seconds. Pretty decent for a 5.5GB model file.

      • kadoban 1 hour ago
        Honestly looks pretty good except whatever is going on with its booty. Is that an ass helmet? I cannot parse what's going on there.
        • bigwheels 55 minutes ago
          I like the lens effect behind the rear tire.
        • Forgeties79 1 hour ago
          I think it’s supposed to be a wing
    • refibrillator 1 hour ago
      Where did you get these instructions?

      They have a demo repo with a setup.sh script:

      https://github.com/PrismML-Eng/Bonsai-demo

      The release tag and weight file you suggest doesn’t match what they wrote.

      • simonw 1 hour ago
        I figured them out, starting from the GGUF on Hugging Face.

        If you have found better instructions and they work then use those instead!

        Personally I prefer to download models directly rather than running some `./setup.sh` script where I need to then review what it does first.

        • refibrillator 1 hour ago
          Yeah just wanted to mention in case it explains the 2x lower throughout you are seeing on M5. To be fair their documentation is a bit inconsistent in some spots.

          Would be good to know if the release and weights from their demo repo work better. I’m trying on a 4090 and will report back.

    • nikwen 1 hour ago
      It would be great to have upstream llama.cpp support for this!
  • adrian17 2 hours ago
    > Ternary Bonsai 2 27B uses ternary {−1, 0, +1} weights with FP16 group-wise scaling, for 1.76 effective bits per weight

    If I recall correctly, a recent post [1] has shown that Q2 quants (with like 2.6 bpw) of the same base Qwen model sit at the edge between "noticeably worse" and Q1's "useless". I took a quick glance at Bonsai's blog posts, and don't really see them comparing themselves to "typical" quants or explaining what's the special sauce that makes them better?

    https://news.ycombinator.com/item?id=49611128

    • nulld3v 52 minutes ago
      There's a table on the HF page that compares it against FP16UD-Q4_K_XL and IQ2_XXS (you need to expand the dropdown): https://huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gguf#fu...

      The table claims it performs on par with UD-Q4_K_XL except on OCR.

    • edflsafoiewq 1 hour ago
      I think the general idea is naive quantization falls apart below 4bpw but you can go lower with more sophisticated QAT-adjacent methods. Bonsai's quantization method is proprietary though.
      • om8 9 minutes ago
        Could've been better if GGUF implemented QTIP format. GGUF representation is a major limitation for llama.cpp quantization performance
        • edflsafoiewq 4 minutes ago
          They use their own llama fork anyway, so that shouldn't matter.
      • yowlingcat 56 minutes ago
        That's correct. I think there can certainly be issues even with 4bpw with naive quantization (IE you'll notice far better results from a QAT 4bpw vs a naive 4bpw).

        One such method that I've been meaning to look into further is Tencent's AngelSlim QAT/PTQ approach. They did a Hy4 preview release thats an STQ_1_0 at 2.38 bpw:

        https://huggingface.co/AngelSlim/Hy4-preview-GGUF https://arxiv.org/abs/2602.21233

        Of course, it's still 213g of VRAM I'd need so it's somewhat out of the range of what I can run locally. In contrast, this new Bonsai is nice because the original was already exciting for making use of low VRAM devices. Could breath new life into some of the older GPUs that were previously close to top of the line just quite VRAM constrained by modern standards and still quite cost effective for now.

    • 0x457 1 hour ago
      1.76 bpw number is kinda misleading if you compare it directly to IQ2/Q2. The encoding is ternary, but the quantization procedure is way more sophisticated than "round Qwen weights to {-1,0,+1}."

      They rotate the weights into a quantization-friendly basis first, then ternarize with per-group scales and error compensation.

  • Aurornis 2 hours ago
    These are small enough that you can run them entirely in the browser https://huggingface.co/spaces/webml-community/ternary-bonsai...

    Remember to clear the downloaded weights afterward.

    Like the last model, it's amazing they work as well as they do. Use it for any longer task and they fall apart spectacularly and in interesting ways.

    • outofpaper 2 hours ago
      So you have some fun examples?
  • danbrooks 2 hours ago
    Nice! Does anyone know how this compares to the Unsloth quantizations of this model? https://unsloth.ai/docs/models/qwen3.8#run-qwen3.8-guide
    • 0xbadcafebee 56 minutes ago
      Came to ask the same. From my really rough understanding, it seems like Unsloth's method allows a slightly higher precision at a higher file size, while PrismML's uses a different approach to achieve a smaller size (and presumably less precision).
  • miffy900 2 hours ago
    I really wish people would stop saying N times smaller than something when making a comparison; that makes no sense - it's 1/9th (11.11%) the size. You don't get a smaller quantity by multiplying by a number greater than 1.0. You could instead reverse the subjects being compared - "the original model is 9x bigger than this new smaller, efficient model" or some such. That makes sense.

    I keep seeing this being used when people talk about efficiency or performance gains and it's just very unintuitive language.

    • zamadatix 1 hour ago
      I agree it makes little sense in a literal mathematical take but "it's 9x smaller" or is too much of linguistic advantage compared to "the original is 9x larger" or "it's 1/9th as large" to expect a change with. You don't have to invoke fractions, it keeps the thing in focus as the first subject, it matches the pattern of the inverse statement, and it's just plain short... so that's what people will adapt and interpret the meaning to be.

      One other way to map both types of linguistic statement consistently to math is to interpret "9x" as "there is a 9 times difference between these two things" and then "smaller"/"larger" tells you which end of that separation the subject is (rather than specifying whether the multiplication builds up or down).

    • hamandcheese 2 hours ago
      If we were talking about speed instead of size, i think it would be perfectly reasonable to say 9x faster. I'm not sure I agree that 9x smaller is unintuitive. It makes sense to me.
      • _carbyau_ 1 hour ago
        "faster" relates to speed. Speed is related to time and speed of a thing is usually defined by time. 9x faster speed translates to time/9. There is an extra step of related conversion there.

        Conversely, 9x [filesize/natural number] is bigger. Every time. At least in the basic maths used by most people. There is no conversion into other units.

        Therefore "9x smaller" when talking about a natural number like filesize is a nonsense statement in logic terms. If you strive for unambiguous phrasing - which is a significant part of the programming experience - this logical nonsense might well perturb you.

        But english language is a flexible thing and if the phrase communicates your intent to your audience then that's fine by me.

      • simondotau 1 hour ago
        "Nine times" literally means multiplied by nine, but here we're dividing by nine. It's not unintelligible (because the corrupted verbiage is so commonplace) but it is needlessly awkward. Like saying "resulted in a size reduction increase of 10 megabytes."
    • rpdillon 15 minutes ago
      Eh, when I read smaller with an integer multiplier, I mentally switch to the reciprocal. Easier than convincing the world not to use "9x smaller". Do you feel the same way about "9x faster"? What you're actually measuring is time, and "faster" is the reciprocal of time, similarly to "smaller" being the reciprocal of size.
    • UI_at_80x24 2 hours ago
      Me too!! It's a huge pet peeve. And it's so hard to get people to see how it's linguistically AND mathematically WRONG.
  • hedora 19 minutes ago
    RAM requirements? My current rule of thumb is “a byte per parameter”, but I doubt this runs in 1/9th that (~ 3GiB).

    Also, perf speedup?

    • jjcm 17 minutes ago
      I'm seeing around 7.9GB of ram, 120 tokens/s on a 6000 pro blackwell.
  • circularfoyers 57 minutes ago
    I wonder how their talks with Apple went. Having this run on the TPU opposed to just the GPU, which drains a significant amount of battery life by comparison, is what I'm really interested in.
  • flutetornado 1 hour ago
    GPT Astra did some benchmarking on the DGX Spark. Speed: 34.38 tokens/sec for generation.

    Seems like we don't have a drafter model yet so it could not test with speculative decoding on. ngram speculative decoding did not help too much either - not enough accepted tokens.

    Smaller size I suppose does not mean better performance in this case - we maybe limited by Spark's low memory bandwidth.

  • kamranjon 3 hours ago
    Love this for the folks with 16gb graphics cards - 3.8 27b has been incredible but not quite runnable on anything less than 32gb - will try loading this up on my 16gb intel b50 and see how it goes - not sure these quants can be accelerated by the XPU cores yet but maybe in time!
    • kadoban 2 hours ago
      You can run the ~4 bit quant(s) on 24gb, if you're not _too_ picky on context size.

      This will hopefully be better, though it'd be a _very_ surprising increase in performace at the size they say. Would love to see more about how it benchmarks.

      • spijdar 2 hours ago
        I run Unsloth's UD-Q4_K_S on 20 GB of VRAM (RX 7900 XT) and I get ~90k tokens of context without quantizing KV cache. With 8-bit quantization, I get about a 134k token context window. That's with only one slot, but for me, it works pretty darn well, with 20-35 tok/s depending on how full that window is.
        • orsorna 34 minutes ago
          7900 XT is a sleeper card. When I initially bought it, it was priced at the lowest wattage per $ per GB VRAM (not normalized for token speeds...) Although I ended up swapping for the XTX because that 4GB means everything in just increasing the context window. At 8bit KV my window is over 200k, and although qwen3.8 loves vomiting out tokens as part of its reasoning chain I trust it enough to get assigned tasks done eventually, which I could not say of any model before its release.
  • JonSchneider 2 hours ago
    I'm hoping they release an 8B v2 based on the Qwen 3.8 series in the near future - that would give us a really powerful model that could be run directly on users phones.
    • verdverm 1 hour ago
      That would require Alibaba releasing a Qwen 3.8 8B first
    • sroussey 2 hours ago
      Yes! And maybe get a hf fused webgpu runner for that model so it’s fast!
  • Havoc 1 hour ago
    Cautiously optimistic. The V1 was noticeably weak on world knowledge but here the 3.8 base model is geared more towards reasoning than world knowledge anyway so might not matter as much
  • jedbrooke 1 hour ago
    Running at about 7-8 tok/s (~60 tok/s prefill) on a Mac Mini M2 16GB.

    So far feels smarter than Bonsai 1 27B, it’s slightly larger than the Q1_0 quant. Super exciting stuff :)

  • cmrdporcupine 55 minutes ago
    What I'd love to see is this done for DS4.1 Flash.

    That would bring it down to the point where it can fit in 128GB on things like the Spark or Strix Halo.

  • abraxas 3 hours ago
    I'm not following the local mdoel scene too closely but this seems quite amazing. Is this able to be run on Apple silicon too?
    • Havoc 3 hours ago
      Their first 27B bonsai was able to run on an iphone.
    • kamranjon 3 hours ago
      "Ternary Bonsai 2 27B reaches up to 143 tokens/second on NVIDIA GeForce RTX 5090 and 46.8 tokens/second on M5 Max. On an RTX 4090, Ternary Bonsai 2 27B consumes just 0.714 mWh/token, making it 40% more energy-efficient than an 8B model running in full-precision."
      • pizza234 2 hours ago
        Their mention of the 5090 is bit odd, since on 32 GB GPUs, Q6 fits while having better quality. Very interesting model for 16 GB GPUs though!
        • sisve 2 hours ago
          They mention 5090 with regards to speed, Q6 will not have that speed?

          And speed matters a lot for many use cases

          • selectodude 1 hour ago
            150 tokens per second on a ternary model implies that it’s GPU bound, I’d bet a Q6 model is even faster because it’s existed longer and seen more optimization. You’d have to be insane to not run an NVFP4 quant over a ternary quant on Blackwell if they both fit.
      • azatom 2 hours ago
        it is like "my fridge is 2mkm (millikilometer) from my desk" m=0.001 h=3600 it should be just Ws or just J
  • 2001zhaozhao 2 hours ago
    I think if they made this for Qwen3.8-Next it could fit in a single 5090?
  • z2 2 hours ago
    I'd love to see a Bonsai model start with a 100B+ parameter model and get that down to <30 GB. But maybe at that point we call it Topiary?
    • all2 1 hour ago
      Other names that occur to me: Orchard, Forest, Stand (of trees).
  • logicallee 1 hour ago
    (In case anyone remembers the compression post from yesterday[1], I checked and this one doesn't qualify for further compression - it's not zero-biased at all.)

    [1] https://news.ycombinator.com/item?id=49732931

  • ipoole_dev0 57 minutes ago
    [flagged]
  • redlimetea 37 minutes ago
    [dead]