Hugging Face's logo

marinone94
/
whisper-tiny-sv

Automatic Speech Recognition
Transformers
whisper
whisper-event
Generated from Trainer
Inference Endpoints
Model card Files Community
1
whisper-tiny-sv / README.md
marinone94's picture
marinone94
add readme only
6d6b33c
preview code
|
raw
history blame
3.12 kB
metadata
language:
  - danish
  - norwegian
  - swedish
license: apache-2.0
tags:
  - whisper-event
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
  - mozilla-foundation/common_voice_11_0
  - mozilla-foundation/common_voice_11_0
  - babelbox/babelbox_voice
  - NbAiLab/NST
  - NbAiLab/NPSC
  - google/fleurs
  - google/fleurs
  - google/fleurs
metrics:
  - wer
model-index:
  - name: Whisper Tiny Nordic
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: mozilla-foundation/common_voice_11_0 nn-NO
          type: mozilla-foundation/common_voice_11_0
          config: null
          split: None
        metrics:
          - name: Wer
            type: wer
            value: 87.65957446808511
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: babelbox/babelbox_voice nst
          type: babelbox/babelbox_voice
        metrics:
          - name: Wer
            type: wer
            value: 87.65957446808511
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: NbAiLab/NST no-distant
          type: NbAiLab/NST
        metrics:
          - name: Wer
            type: wer
            value: 87.65957446808511
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: NbAiLab/NPSC 16K_mp3_nynorsk
          type: NbAiLab/NPSC
        metrics:
          - name: Wer
            type: wer
            value: 87.65957446808511
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: google/fleurs nb_no
          type: google/fleurs
        metrics:
          - name: Wer
            type: wer
            value: 87.65957446808511

Whisper Tiny Nordic

This model is a fine-tuned version of openai/whisper-tiny on the mozilla-foundation/common_voice_11_0 sv-SE, the mozilla-foundation/common_voice_11_0 da, the mozilla-foundation/common_voice_11_0 nn-NO, the babelbox/babelbox_voice nst, the NbAiLab/NST no-distant, the NbAiLab/NPSC 16K_mp3_nynorsk, the google/fleurs sv_se, the google/fleurs da_dk and the google/fleurs nb_no datasets. It achieves the following results on the evaluation set:

  • Loss: 5.1226
  • Wer: 87.6596

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • training_steps: 1
  • mixed_precision_training: Native AMP

Training results

Framework versions

  • Transformers 4.26.0.dev0
  • Pytorch 1.13.1+cu117
  • Datasets 2.7.1.dev0
  • Tokenizers 0.13.2