Skip to content
Computer vision

Huihui-gemma-4-12B-it-abliterated

Gemma 4 12B без отказов

1 min readupdated

Это uncensored-версия google/gemma-4-12B-it, полученная через abliteration — грубую proof-of-concept реализацию удаления отказов без TransformerLens. Модель на 11,96 млрд параметров, задача any-to-any, библиотека transformers.

Ключевые особенности

  • Abliterated оба режима — thinking и non-thinking.
  • Затронуты только слои 23–28.
  • Есть готовый вариант для ollama: huihui_ai/gemma-4-abliterated:12b.
  • Safety-фильтрация снижена: возможны чувствительные и спорные ответы.
  • Рекомендована для исследований и контролируемых сред, не для продакшена.

Пример

ollama run huihui_ai/gemma-4-abliterated:12b
ПодробнееБенчмарков в карточке нет. Размер — 11,96 млрд параметров, база — google/gemma-4-12B-it.
Abliterated только слои 23–28; оба режима — thinking и non-thinking.
Веса раннего релиза имели проблемы — авторы перезалили модель; если скачали, обновите.
Ограничения: сниженная фильтрация, возможны спорные ответы; для research и контролируемых тестов.
Discuss Telegram 0 22

Will it run?

Runs in FP8 / INT8

Full precision does not fit, a quantised version does: quality barely changes down to 8 bits, more noticeably at 4 bits.

  • BF16weights 22 GB · ready files25 GB
  • FP8 / INT8recommendedweights 11 GB13 GB
  • 4-bitweights 6.3 GB7.4 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 5080 · BF16 — from RTX 5090

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

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.

    # agenz-ia/Huihui-gemma-4-12B-it-abliterated
    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 agenz-ia/Huihui-gemma-4-12B-it-abliterated --local-dir ~/models/Huihui-gemma-4-12B-it-abliterated
    # 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

  1. gemma-4-12Bfamily root
  2. gemma-4-12B-it
  3. Huihui-gemma-4-12B-it-abliterated

New posts about versions of gemma-4-12B (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 4All on Hugging Face

About the model

agenz-ia/Huihui-gemma-4-12B-it-abliterated
Author
agenz-ia
Category
Computer vision
Released
Posted

On NeyroLearn

Hugging Face

220

Ask AI about this model

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

Comments