whisper-small-ewe-2 / README.md
abiyo27's picture
End of training
8f08cf7 verified
|
raw
history blame
1.97 kB
metadata
language:
  - multilingual
license: apache-2.0
base_model: openai/whisper-medium
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - abiyo27/BibleTTS_Ewe-Bible
metrics:
  - wer
model-index:
  - name: Whisper_Small_Ewe
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: BibleTTS
          type: abiyo27/BibleTTS_Ewe-Bible
          config: default
          split: None
          args: 'config: ewe, split: train'
        metrics:
          - name: Wer
            type: wer
            value: 10.094952523738131

Whisper_Small_Ewe

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

  • Loss: 0.1021
  • Wer: 10.0950

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: 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: 14000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.2196 0.1802 4000 0.1780 19.3903
0.1587 0.3605 8000 0.1375 13.4933
0.1162 0.5407 12000 0.1021 10.0950

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.3.1+cu121
  • Datasets 2.19.2
  • Tokenizers 0.19.1