Skip to content
Audio

Whisper tiny Japanese English

компактный ASR-файнтюн с открытыми метриками

1 min read

Это файнтюн openai/whisper-tiny для задачи automatic-speech-recognition от автора merkelmarrow. Модель содержит около 0.04 млрд параметров и обучена на Japanese English dataset; карточка во многом автогенерирована, поэтому описание назначения и ограничений пока не заполнено.

Ключевые особенности

  • База — openai/whisper-tiny, очень компактная ASR-модель.
  • Задача — automatic-speech-recognition, библиотека transformers.
  • На evaluation set указаны loss 0.4743 и WER 25.6736.
  • Обучение: 2000 шагов, train_batch_size 2, learning_rate 1e-5, warmup 500.
  • В карточке нет разделов Intended uses & limitations и Training and evaluation data — перед использованием нужна своя проверка.

Пример

Сценарий: подать короткую аудиозапись с японско-английской речью в pipeline automatic-speech-recognition и сравнить транскрипцию с эталоном. При WER около 25.7 модель стоит рассматривать как эксперимент, а не готовый продакшн-инструмент.

ПодробнееWER на evaluation set: 25.6736; loss: 0.4743.
Обучение: 2000 шагов, train_batch_size 2, learning_rate 1e-5, warmup 500.
Карточка автогенерирована: Model description, Intended uses & limitations, Training and evaluation data не заполнены.
Метаданные указывают язык en; название и датасет — Japanese English. Требования к VRAM не указаны.
Discuss Telegram 0 24

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.

    # merkelmarrow/whisper-tiny-japanese_english-ft1
    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 merkelmarrow/whisper-tiny-japanese_english-ft1 --local-dir ~/models/whisper-tiny-japanese_english-ft1
    # 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 whisper-tiny-japanese_english-ft1 (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

merkelmarrow/whisper-tiny-japanese_english-ft1
Author
merkelmarrow
Category
Audio
Released
Posted

On NeyroLearn

Hugging Face

240

Ask AI about this model

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

Comments