loom-fun/LiquidAI-d1-2B-loom

d1-2B-GGUF for AMD Loom

A Loom package (.loom programs and a model.json card) for the unmodified GGUF checkpoint LiquidAI/d1-2B-GGUF (Q4_K_M), generated on an AMD Ryzen AI Max+ 395 "Strix Halo" by the loom-skills factory workflow.

The weights are not included. loom-fetch (from the Loom SDK, installed with its hf extra by the first command) installs the programs from this repository and the GGUF from LiquidAI/d1-2B-GGUF, checked against model.json.lock:

pip install "hrx-loom[hf]"
loom-fetch https://huggingface.co/loom-fun/LiquidAI-d1-2B-loom/resolve/main/model.json --materialize=copy --output=d1-2B-GGUF
loom-serve --model=d1-2B-GGUF/model.json --device=amdgpu

This is a decision model, not a chat model: it answers choice, score and noul questions about a text or JSON state in one forward pass, without generating text. The server listens on 127.0.0.1:8080 and answers them at /v1/systemone:

curl -s http://127.0.0.1:8080/v1/systemone -H 'Content-Type: application/json' -d '{
  "model": "LiquidAI/d1-2B-GGUF",
  "state": "I was charged twice this month, please refund one of them.",
  "questions": {
    "refund": {
      "type": "noul",
      "instructions": "Is the customer asking for a refund?"
    },
    "team": {
      "type": "choice",
      "instructions": "Which team should handle this?",
      "criteria": {
        "billing": "Charges, refunds, invoices",
        "technical": "App or site faults"
      }
    }
  }
}'

Choice and score answers come back with their full probability maps, and noul answers with the probability of true.

LiquidAI/d1-2B-GGUF is tagged image-text-to-text, but this Loom package takes text input only.