Skip to content
Computer vision

OneJev1.5-4B

мультимодальная модель решений

1 min read

OneJev1.5-4B от команды OmniJev — 4B-модель для принятия решений. Это полный fine-tune Qwen/Qwen3.5-4B, обученный примерно на 320 тыс. вопросов из реальных agent runs, видео, изображений и тренировочных частей публичных decision-датасетов. На вход принимает скриншот, фото, видео или текст с вопросами, а на выходе за один forward pass даёт калиброванную вероятность для каждого варианта.

Ключевые особенности

  • Мультимодальный вход: изображения, видео и обычный текст.
  • Один forward pass — вероятности по всем вариантам.
  • Калиброванные вероятности для вариантов ответа или действия.
  • API в стиле TypeSafe System One с полем media для изображений и видео.
  • Обучение на ~320 тыс. вопросов из реальных прогонов агентов и публичных датасетов.
  • Лицензия Apache 2.0.

Пример

from qev import Client, Noul
from qev.media import data_uri

r = Client("http://localhost:8000").system_one(
    state={"task": "Pay the open invoice from ACME", "screen": "<image:1>"},
    media=[{"type": "image", "data": data_uri("screenshot.png")}],
    questions={"done": Noul("The invoice has been paid")},
)
ПодробнееDecision Index 0.3, публичная часть: 54.89 — второй результат среди 83 моделей до 8B, отставание от лидера 0.01.
Обучение: полный fine-tune Qwen/Qwen3.5-4B на ~320 тыс. вопросов.
Лицензия Apache 2.0; модель выдаёт вероятности вариантов за один forward pass.
Требования к VRAM в карточке не указаны.
Discuss Telegram 0 38

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.

    # qhoenix/zils-mini-onejev15
    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 qhoenix/zils-mini-onejev15 --local-dir ~/models/zils-mini-onejev15
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to zils-mini-onejev15
  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 zils-mini-onejev15 (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

qhoenix/zils-mini-onejev15
Author
qhoenix
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

380

Ask AI about this model

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

Comments