whisper-small-hi / README.md
dussen's picture
End of training
1c6a66c
metadata
language:
  - ita
license: apache-2.0
base_model: openai/whisper-small
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: Whisper Small - Italian
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0
          type: mozilla-foundation/common_voice_11_0
          config: it
          split: test
          args: 'config: Italian, split:  test'
        metrics:
          - name: Wer
            type: wer
            value: 48.7527114967462

Whisper Small - Italian

This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6783
  • Wer: 48.7527

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: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 50
  • training_steps: 200
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.024 0.79 50 0.6568 53.1453
0.0154 1.59 100 0.6739 41.8655
0.0146 2.38 150 0.6809 41.6486
0.0097 3.17 200 0.6783 48.7527

Framework versions

  • Transformers 4.36.0.dev0
  • Pytorch 2.1.1+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0