Skip to content
Other

Chon-1.0-12B-Fast

быстрые typed-decision решения одним forward pass

2 min read

Chon-1.0-12B-Fast от ninhnv — typed-decision модель: на вход получает state (текст, JSON или диалог) и один вопрос с закрытым набором вариантов, а на выходе даёт калиброванные вероятности для каждого варианта. Это LoRA-адаптер (~1.0 GB) на google/gemma-4-12B-it (12B); поддерживаются английский и вьетнамский.

Ключевые особенности

  • choice: 2–256 вариантов; больше 16 читаются группами, затем финальный раунд.
  • yes/no и score: P(yes) или распределение по упорядоченным уровням.
  • Fast-вариант: reasoning отключён, каждое решение — один forward pass, output_tokens = 0.
  • Возвращает вероятности по всем вариантам, а не только выбранную метку.
  • Вход: state до 6 000 токенов, prompt до 8 192.
  • Лицензия Apache-2.0, как у базовой модели.

Пример

from chon.infer import DecisionModel

dm = DecisionModel.load("ninhnv/chon-1.0-12b-fast", device="auto")
answers, usage = dm.predict(
    "Order #4411 shipped on 3 March; the customer reports it arrived damaged on 9 March.",
    {"damaged": {"type": "noul", "instructions": "Did the order arrive damaged?"}})
print(answers["damaged"]["noul"])
ПодробнееБенчмарки: held-out replica (720) — Intelligence 72.4, Calibration 86.5, Capability 79.4; JevBench public (231) — 78.8 / 87.8 / 83.3.
Железо: одна GPU от 40 GB для bf16 (замер на H100 80 GB).
Скорость: медиана 0.07 с, p95 0.1 с; 0 output tokens на решение.
Ограничения: state до 6 000 токенов, prompt до 8 192; языки — английский и вьетнамский.
Discuss Telegram 0 92

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # ninhnv/chon-1.0-12b-fast
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download ninhnv/chon-1.0-12b-fast --local-dir ~/models/chon-1.0-12b-fast
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to chon-1.0-12b-fast
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of chon-1.0-12b-fast (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

ninhnv/chon-1.0-12b-fast
Author
ninhnv
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

920

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments