numidlabs / wazn-2b-v0.1 Like 1 Follow numidlabs 2

wazn-2b-v0.1

Wazn is a decision model: given a context, optional rules and an instruction, it returns a probability distribution over a set of labels you supply, each with an optional definition and few-shot examples. Nothing is generated.

Experimental. For trying the model out, not for production.

How it works

1. Encode each candidate. The rules, context and instruction are followed by one candidate label (its definition and examples), closed by a `</choice>` tag. The frozen Qwen3.5-2B backbone, with LoRA adapters, reads this sequence, and the hidden state at `</choice>` is projected into a representation of that candidate. 2. Compare the candidates. A relational judge compares the candidate representations against each other, rather than scoring each label alone, and outputs a probability for each one. 3. Check that any label fits. A NONE gate estimates the probability that at least one candidate is valid. Below a threshold the answer is NONE; otherwise it is the most likely candidate.

Use

Python 3.11+. Install the wazn-experimentalpackage in a virtual environment, then start a local server with this checkpoint:

python -m venv .venv && source .venv/bin/activate
pip install --pre "wazn-experimental[server]"
wazn-experimental serve --model numidlabs/wazn-2b-v0.1
serving on http://127.0.0.1:8000

• `--pre` is needed while the package only has alpha releases (`0.1.0aN`).
• Keep the quotes around `"wazn-experimental[server]"`; zsh treats the brackets as a glob.
• `[server]` installs PyTorch (a large download). On Linux with an NVIDIA GPU, use `"wazn-experimental[server,cuda]"` for the fast kernels.

Then query it from Python:

from wazn_experimental import Client, Request, Instruction, Label
request = Request(
    context="My running shoes arrived in the wrong size. Can I swap them?",
    instructions=[Instruction(
        "Which team should handle this?",
        labels=[Label("returns", "Exchanges, refunds, wrong or damaged items"),
                Label("shipping", "Delivery status, delays, lost packages"),
                Label("billing", "Charges, invoices, payment problems")],
    )],
)
print(Client().predict(request).answer)

Model

• Backbone: `Qwen/Qwen3.5-2B-Base` at revision `b1485b2fa6dfa1287294f269f5fb618e03d52d7c`, loaded from the hub (not in this repo)
• This repo: the head (43,214,082 parameters: relational judge, NONE gate, tag embeddings, LoRA adapters)

Limitations

TODO: known weak spots, training data summary, evaluation results.

License

The head weights are released under Apache-2.0. The backbone keeps its own license; see `Qwen/Qwen3.5-2B-Base`.

Downloads last month 37

Inference Providers NEW

Zero-Shot Classification

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for numidlabs/wazn-2b-v0.1 [](https://huggingface.co/docs/hub/model-cards#specifying-a-base-model)

Base model

Qwen/Qwen3.5-2B-Base

Finetuned

(97)

this model