whisper-tiny-am / README.md
Bedru's picture
End of training
4c8ea04 verified
metadata
library_name: transformers
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - generated_from_trainer
datasets:
  - common_voice_17_0
metrics:
  - wer
model-index:
  - name: whisper-tiny-am
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: common_voice_17_0
          type: common_voice_17_0
          config: am
          split: None
          args: am
        metrics:
          - name: Wer
            type: wer
            value: 212.75698471270425

whisper-tiny-am

This model is a fine-tuned version of openai/whisper-tiny on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.6214
  • Wer: 212.7570

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 150
  • training_steps: 100
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
3.047 0.5682 25 2.7261 294.8867
2.3761 1.1364 50 2.1202 412.2298
1.9305 1.7045 75 1.7593 104.0063
1.6908 2.2727 100 1.6214 212.7570

Framework versions

  • Transformers 4.48.3
  • Pytorch 2.5.1+cu124
  • Datasets 3.3.0
  • Tokenizers 0.21.0