Skip to content
Computer vision

LFM3-2.2B-Omni-Coder-Instruct

компактный омни-модальный кодер

1 min read

Кастомный merge от Shivansh Kokate: бэкбон LFM3-1.7B объединён с Vision, Audio, Image и Video-компонентами LFM3-Omni-2B. Получилась instruct-модель на 2,24 млрд параметров для кода, tool calling и мультимодального понимания на edge-устройствах.

Ключевые особенности

  • Продвинутое кодирование: генерация, отладка и объяснение кода в небольшом размере.
  • Tool calling: структурированные вызовы функций и работа с внешними API.
  • VIA-понимание: изображения, видео со звуком или без и аудио 16 kHz вместе с текстом.
  • Контекст 32K токенов на вход и выход.
  • Выходы: текст, tool calls и код.

Пример

Сценарий из карточки: подайте текст, изображение, видео со звуком или без и аудио 16 kHz — модель вернёт текст, tool calls или код. Для запуска нужны Transformers, torch, torchvision, accelerate, pillow, librosa и soundfile.

ПодробнееПараметры: 2 240 487 474 (2,24B). Контекст: 32K токенов.
Входы: текст, изображения, видео со звуком/без и аудио 16 kHz; выходы: текст, tool calls, код.
Формальных бенчмарков для этого merge пока нет — заявлены унаследованные способности базовых LFM3.
Лицензия LFM Open License v1.0; требования к VRAM в карточке не указаны. Скоро обещаны GGUF, первым Q4_K_M.
Discuss Telegram 1 52

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.

    # Shivansh-13929/LFM3-2.2B-Omni-Coder-Instruct
    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 Shivansh-13929/LFM3-2.2B-Omni-Coder-Instruct --local-dir ~/models/LFM3-2.2B-Omni-Coder-Instruct
    # The model is in ~/models. How to run it is in the model card (link above).
  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 LFM3-2.2B-Omni-Coder-Instruct (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

Shivansh-13929/LFM3-2.2B-Omni-Coder-Instruct
Author
Shivansh-13929
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

521

Ask AI about this model

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

Comments

Computer vision

Mage Hand

Mage Hand
Computer vision

Viper-9B

Viper-9B
Computer vision

PANORAMA-2B

PANORAMA-2B