README.md

metadata

library_name: transformers
language:
  - en
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - generated_from_trainer
datasets:
  - wwwtwwwt/fineaudio-Education
metrics:
  - wer
model-index:
  - name: Whisper Tiny En - Education - Documentaries
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: fineaudio-Education-Documentaries
          type: wwwtwwwt/fineaudio-Education
          args: 'config: en, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 46.58774235626618

Whisper Tiny En - Education - Documentaries

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

• Loss: 1.4409
• Wer: 46.5877

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 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.7835	0.8230	1000	1.3163	54.4973
0.5343	1.6461	2000	1.3817	50.0405
0.4552	2.4691	3000	1.4197	46.4158
0.3277	3.2922	4000	1.4409	46.5877

Framework versions

• Transformers 4.46.3
• Pytorch 2.4.0
• Datasets 3.1.0
• Tokenizers 0.20.0