whisper-tiny-en-US / README.md
ptah23's picture
update model card README.md
d81a0f9
|
raw
history blame
2.3 kB
metadata
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - generated_from_trainer
datasets:
  - PolyAI/minds14
metrics:
  - wer
model-index:
  - name: whisper-tiny-en-US
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: PolyAI/minds14
          type: PolyAI/minds14
          config: en-AU
          split: train
          args: en-AU
        metrics:
          - name: Wer
            type: wer
            value: 0.20146619603584034

whisper-tiny-en-US

This model is a fine-tuned version of openai/whisper-tiny on the PolyAI/minds14 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6756
  • Wer Ortho: 0.2044
  • Wer: 0.2015

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: 16
  • 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: 4000

Training results

Training Loss Epoch Step Validation Loss Wer Ortho Wer
0.0007 17.86 500 0.5138 0.1941 0.1920
0.0002 35.71 1000 0.5565 0.1958 0.1936
0.0001 53.57 1500 0.5851 0.1981 0.1958
0.0001 71.43 2000 0.6081 0.2030 0.1998
0.0 89.29 2500 0.6273 0.2038 0.2009
0.0 107.14 3000 0.6441 0.2021 0.1996
0.0 125.0 3500 0.6602 0.2035 0.2007
0.0 142.86 4000 0.6756 0.2044 0.2015

Framework versions

  • Transformers 4.31.0.dev0
  • Pytorch 1.12.1+cu116
  • Datasets 2.4.0
  • Tokenizers 0.12.1