whisper-small-bg / README.md
BKat's picture
End of training
f054234
metadata
language:
  - bg
license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_13_0
metrics:
  - wer
model-index:
  - name: openai/whisper-small-finetuned-common_voice_13_0-bg
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: mozilla-foundation/common_voice_13_0
          type: mozilla-foundation/common_voice_13_0
          config: bg
          split: test
          args: bg
        metrics:
          - name: Wer
            type: wer
            value: 23.264792642720806

openai/whisper-small-finetuned-common_voice_13_0-bg

This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_13_0 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3983
  • Wer Ortho: 30.2504
  • Wer: 23.2648

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: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant_with_warmup
  • lr_scheduler_warmup_steps: 50
  • training_steps: 1000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer Ortho Wer
0.0787 2.78 500 0.3445 31.2999 24.2365
0.0145 5.56 1000 0.3983 30.2504 23.2648

Framework versions

  • Transformers 4.36.2
  • Pytorch 1.12.0+cu102
  • Datasets 2.15.0
  • Tokenizers 0.15.0