whisper-small-zh-tw / README.md
kimbochen's picture
update model card README.md
e393db1
|
raw
history blame
2.09 kB
metadata
language:
  - zh
license: apache-2.0
tags:
  - whisper-event
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: Whisper Small Chinese - Kimbo Chen
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0
          type: mozilla-foundation/common_voice_11_0
          config: zh-TW
          split: test
          args: zh-TW
        metrics:
          - name: Wer
            type: wer
            value: 42.517911975435005

Whisper Small Chinese - Kimbo Chen

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: 0.2554
  • Wer: 42.5179

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: 64
  • 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: 500
  • training_steps: 5000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.0046 6.02 1000 0.2102 41.8014
0.0014 13.02 2000 0.2266 41.6172
0.0004 20.01 3000 0.2437 42.1699
0.0003 27.0 4000 0.2521 42.5179
0.0002 33.02 5000 0.2554 42.5179

Framework versions

  • Transformers 4.26.0.dev0
  • Pytorch 1.13.0+cu116
  • Datasets 2.7.1.dev0
  • Tokenizers 0.13.2