Skip to content
Audio

Pianissimo-da

открытый ASR для датского с WER 11,8%

1 min read

Klang AI выпустила открытую модель распознавания датской речи Pianissimo-da на базе NVIDIA Parakeet v3. Это FastConformer + TDT decoder примерно на 600 млн параметров; поддерживаются пунктуация, регистр и таймстемпы.

Ключевые особенности

  • Средний WER 11,8% на трёх датских наборах — лучший результат среди протестированных открытых и закрытых моделей.
  • Более 3 600× realtime на одной NVIDIA H100 и 542× на L4.
  • Поддерживает пунктуацию, регистр и таймстемпы на уровне слов.
  • Есть версии для серверов, ноутбуков и телефонов: ONNX, MLX, GGUF и CoreML.
  • Лицензия CC BY 4.0 — можно запускать на своём железе.

Пример

import torch
from nemo.collections.asr.models import ASRModel

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = ASRModel.from_pretrained(model_name="KlangAI/pianissimo-da", map_location=device).eval()
with torch.inference_mode():
    hyp = model.transcribe(audio=["audio.wav"], batch_size=1, return_hypotheses=True)
print(hyp[0].text)
ПодробнееWER: 11,8% в среднем; EU Parliament — 7,9%, DR TV — 18,2%, DanPASS — 9,4%.
Обучение: около 70 000 часов датской речи, 470 000 шагов.
Чекпойнт fp32 — 2,51 ГБ; квантованные версии — от 485 МБ (GGUF Q4_K_M).
Ограничение: модель только для датского языка.
Discuss Telegram 1 16

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.

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

Quantised versions 5All on Hugging Face
Fine-tunes 2All on Hugging Face

About the model

KlangAI/pianissimo-da
Author
KlangAI
Category
Audio
Released
Posted

On NeyroLearn

Hugging Face

161

Ask AI about this model

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

Comments

Audio

OmniVoice

OmniVoice
Audio

AMBIT

AMBIT
Audio

StaleyNET

StaleyNET