any2jev-qwen3-0.6b-snake

A Jev-style System One decision model made with any2jev from Qwen/Qwen3-0.6B. State in, typed Choice / Score / Noul answers with calibrated probabilities out, in one forward pass. No text is generated.

pip install "any2jev[serve]"
any2jev serve hf://huaweifeng/any2jev-qwen3-0.6b-snake      # POST /v1/systemone, TypeSafe SDK compatible
any2jev ask hf://huaweifeng/any2jev-qwen3-0.6b-snake --state "My payouts have failed 3 days in a row, fix this ASAP" \
    --choice "Which team? | billing, technical, sales" --noul "Is this urgent?"

What is in this repo

• adapter/: LoRA adapter (r=16) on Qwen/Qwen3-0.6B, trained with the vocabulary head removed
• head.safetensors: the pointer head (dim 256) that scores options against the decision token
• any2jev.json: delimiters, mode (packed), temperature 1.509 fitted on validation
• tokenizer/: the base tokenizer (delimiter tokens reused or added)
• train_report.json, eval.json: training config, history and held-out metrics

Training

• data: data/snake/train.jsonl; 10.6 M trainable parameters, 1.0 epoch(s), lr 0.0002, batch 8 x 1
• wall clock: 13 min on one consumer GPU

Held-out evaluation

group	n	accuracy	NLL	Brier	ECE	AURC
overall	400	0.953	0.191	0.087	0.034	0.019
choice	400	0.953	0.191	0.087	0.034	0.019
Isolation check (packed vs separate questions): max	dp	= 0.0e+00.

Limitations

Trained on a few thousand labelled decisions from a handful of sources; expect the accuracy above on similar inputs and lower accuracy off-distribution. Probabilities are calibrated on the validation split, not a guarantee per answer. Keep arithmetic, dates and counting in code, as Jev's own docs recommend.

Independent project, not affiliated with TypeSafe AI. Apache-2.0.