Skip to main content

Quyet

Quyet is a family of decision models. Give it a state (any text, JSON or conversation) and one or more typed questions; it picks one option per question and returns calibrated probabilities. Nothing is generated: every answer is a probability distribution over the options you supplied.

This package is the runtime for all Quyet 1.0 models. The weights are on Hugging Face.

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

The models are built for English and also tuned for Vietnamese; other languages work, with lower accuracy. Small-EN is trained on English only.

model kind parameters
Quyet-1.0-Large LLM (GPU) 31.3B (30.7B text)
Quyet-1.0-Medium LLM (GPU) 4.66B (4.21B text)
Quyet-1.0-Small encoder (CPU or GPU) 328M
Quyet-1.0-Small-EN encoder (CPU or GPU) 153M
Quyet-1.0-Tiny encoder (CPU or GPU) 183M

Install

pip install quyet                 # CPU or single GPU
pip install "quyet[multi-gpu]"    # spread Quyet-1.0-Large over several GPUs (device_map="auto")

Python 3.10 or newer. The first quyet.load downloads the weights from Hugging Face.

Use

import quyet

m = quyet.load("chinhnc/Quyet-1.0-Small")
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"])

quyet.load(name_or_path, device=None, revision=None, dtype=None, device_map=None) accepts a Hugging Face repo id or a local directory. predict_many answers a list of requests; raw_probabilities returns the uncalibrated distributions.

Question types and answers

Answers follow the TypeSafe /v1/systemone shape:

  • choice: pick one label from criteria (a dict of label to description, or a list of labels). Answer: choice, confidence, probabilities.
  • score: an ordered scale given as a list of levels. Answer: expected level score, probabilities, legend.
  • noul: a statement that is true or false. Answer: noul = P(true).

Probabilities are temperature-calibrated per question type and option count.

Command line

quyet predict chinhnc/Quyet-1.0-Small < requests.jsonl > answers.jsonl

Each input line is {"state": ..., "questions": {...}}; each output line is the response or {"error": ...}. Options: --input FILE, --device cpu|cuda|cuda:1, --strict (fail instead of truncating a long state).

Limits

  • At most 10 options per question; more are rejected, not truncated.
  • Only the state is ever truncated: conversation lists keep their most recent turns, other states keep their beginning. Encoders read 8,192 tokens; the LLMs keep up to 6,000 state tokens.
  • Not evaluated for safety, bias or adversarial robustness; do not use as the only control for high-stakes decisions.

Licence and credit

Apache-2.0 for the code and the weights. Keep the NOTICE file, which starts with "Quyet by Chinh Nguyen", when you redistribute this package, a model, or anything derived from them.

@misc{quyet2026,
  title  = {Quyet 1.0: calibrated decision models},
  author = {Chinh Nguyen},
  year   = {2026},
  url    = {https://github.com/ncchinh/quyet}
}

Questions and issues: GitHub issues or email@chinh.com.

Metadata

Release files for quyet 1.0.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for quyet 1.0.2
File Size Uploaded Provenance Actions
quyet-1.0.2.tar.gz 26.1 kB Without provenance Details for quyet-1.0.2.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for quyet 1.0.2
File Interpreter ABI Platform Actions
quyet-1.0.2-py3-none-any.whl Python 3 none any Details for quyet-1.0.2-py3-none-any.whl

Total release size: 51.1 kB

Release files / quyet-1.0.2.tar.gz

Download URL quyet-1.0.2.tar.gz
Size 26.1 kB
Tags Source
SHA-256 checksum
How to use checksums
66a3c2f01f6c0dc30be0b343cbfafd130682627852c550e9185ec54a4e19174b
BLAKE2b-256 checksum
How to use checksums
e0471d70d0ad83eef8f81766bf8ef0ea786c60004ab6662c1c48a40943683762
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.12.10 {"installer":{"name":"uv","version":"0.12.10","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Release files / quyet-1.0.2-py3-none-any.whl

Download URL quyet-1.0.2-py3-none-any.whl
Size 25.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
60a36fdb575aec9f1b82462856ebacd165a9d2941cf68b3b83c044ecdec9a1d0
BLAKE2b-256 checksum
How to use checksums
3a475b9117bb8ecd92365bf6cfaab44ee2467f3803f6fe380ae1ac3e196dfb75
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.12.10 {"installer":{"name":"uv","version":"0.12.10","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Security for quyet 1.0.2

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page