Skip to content
Audio

NB-Whisper medium Core ML

норвежская речь локально на Apple Neural Engine

1 min read

Это конвертация NbAiLab/nb-whisper-medium от Национальной библиотеки Норвегии в Core ML для WhisperKit, а не новая модель. Веса и поведение — оригинальные от NbAiLab; формат позволяет запускать распознавание норвежской речи на Apple Neural Engine. Конверсия используется в приложении Sagt для Mac, iPhone и iPad.

Ключевые особенности

  • Core ML-формат для WhisperKit: запуск на Apple Neural Engine.
  • Оригинальные веса и поведение NbAiLab, лицензия Apache 2.0.
  • medium по точности сравним с NB-Whisper large на замерах, но вдвое меньше.
  • Поддерживает норвежский: no, nb, nn; tokenizer как у openai/whisper-medium.
  • Конверсия в float16 без квантования; config.json и generation_config.json скопированы без изменений.

Пример

let folder = try await WhisperKit.download(
    variant: "NbAiLab_nb-whisper-medium",
    from: "Selbyggen/nb-whisper-medium-coreml"
)
let kit = try await WhisperKit(WhisperKitConfig(modelFolder: folder.path))
let result = try await kit.transcribe(audioPath: "opptak.m4a",
                                      decodeOptions: DecodingOptions(language: "no"))
ПодробнееWER на реальной норвежской речи: FLEURS nb — 6.5% (large 7.0%), NST — 2.6% (large 2.2%), спонтанная NB Tale — 15.8% (large 17.1%).
На Neural Engine medium показал точность уровня large при вдвое меньшем размере.
Ограничения: только норвежский язык и формат Core ML/WhisperKit; это конверсия, а не новая модель.
Discuss Telegram 0 16

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 0.9 GB · ready files1.4 GB
  • FP8 / INT8recommendedweights 0.4 GB1.0 GB
  • 4-bitweights 0.2 GB0.8 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 2060 · BF16 — from RTX 2060

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.

    # Selbyggen/nb-whisper-medium-coreml
    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 Selbyggen/nb-whisper-medium-coreml --local-dir ~/models/nb-whisper-medium-coreml
    # 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. whisper-mediumfamily root
  2. nb-whisper-medium
  3. nb-whisper-medium-coreml

New posts about versions of whisper-medium (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 6All on Hugging Face

About the model

Selbyggen/nb-whisper-medium-coreml
Author
Selbyggen
Category
Audio
Released
Posted

On NeyroLearn

Hugging Face

160

Ask AI about this model

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

Comments