metadata

language:
  - en
license: apache-2.0
base_model: openai/whisper-small
tags:
  - domain-asr
  - generated_from_trainer
datasets:
  - audiofolder
metrics:
  - wer
model-index:
  - name: Whisper
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: immunology dataset
          type: audiofolder
          config: default
          split: test
          args: 'config: en, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 9.337797619047619

Whisper

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

Loss: 0.3058
Wer: 9.3378

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
training_steps: 4000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0259	4.55	1000	0.2254	9.5610
0.0135	9.09	2000	0.2853	9.375
0.0022	13.64	3000	0.2989	9.375
0.0004	18.18	4000	0.3058	9.3378

Framework versions

Transformers 4.39.0.dev0
Pytorch 2.1.0+cu121
Datasets 2.17.1
Tokenizers 0.15.2