Quyet-1.0-Medium

Quyet-1.0-Medium is a decision model: given a state (any text, JSON or conversation) and one or more typed questions, it picks one option per question and returns calibrated probabilities. Question types: choice (pick one label), score (an ordered scale) and noul (true / false). It is part of the Quyet 1.0 family (Large, Medium, Small, Small-EN, Tiny), released by Chinh Nguyen under Apache-2.0.

Base model	Qwen/Qwen3.5-4B
Architecture	Qwen3.5-4B with a merged LoRA fine-tune (rank 16), letter-readout decision prompt
Parameters	4.66B (4.21B text)
Languages	English; also tuned for Vietnamese. Other languages work, with lower accuracy.
Input	state up to 6,000 tokens inside an 8,000-token prompt
Weights	9.3 GB (bf16)
License	Apache-2.0 (see LICENSE and NOTICE)
Homepage	quyet.ai

How to use

Live demo: try Quyet-1.0-Large in your browser at quyet.ai.

pip install quyet

import quyet

m = quyet.load("chinhnc/Quyet-1.0-Medium")          # pip install quyet; downloads from Hugging Face
r = m.predict(
    {"message": "Please close my card, I lost it yesterday."},
    {"intent": {"type": "choice", "instructions": "What does the customer want?",
                 "criteria": {"cancel": "close the card", "limit": "change the limit", "other": None}},
     "urgent": {"type": "noul", "instructions": "The request is urgent."},
     "mood": {"type": "score", "instructions": "How upset is the customer?", "criteria": ["calm", "annoyed", "angry"]}},
)
print(r["answers"])   # {"intent": {"choice": ..., "confidence": ..., "probabilities": {...}}, "urgent": {"noul": P(true)}, ...}

Runs in bf16 on a 16 GB GPU. The model answers by reading the next-token probabilities of the option letters (A, B, ...) after a fixed prompt; the quyet package builds that prompt and applies the calibrated temperatures. It also loads with transformers (and serves with vLLM) as a standard Qwen3.5-4B-architecture checkpoint, but the decision prompt and temperatures live in the package and in quyet_config.json. This model uses prompt version 1, the prompt it was trained with.

Answers follow the TypeSafe /v1/systemone shape: choice (with probabilities), score (expected level, probabilities, legend) and noul (P(true)). At most 10 options per question. Only the state is ever truncated: conversation lists keep their most recent turns, other states keep their beginning.

Credits

• Qwen3.5 by Alibaba Cloud (Apache-2.0).

Citation

@misc{quyet2026,
  title  = {Quyet 1.0: calibrated decision models},
  author = {Chinh Nguyen},
  year   = {2026},
  url    = {https://huggingface.co/chinhnc/Quyet-1.0-Medium}
}

License

Apache-2.0. Keep the NOTICE file (it starts with "Quyet by Chinh Nguyen") when you redistribute this model or anything derived from it. Questions and issues: email@quyet.ai.