README.md

metadata

language:
  - ar
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - generated_from_trainer
datasets:
  - ubulut/quran-verses
metrics:
  - wer
model-index:
  - name: Whisper Tiny AR - Quran
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: quran-whisper-dataset
          type: ubulut/quran-verses
          config: default
          split: None
          args: 'config: ar, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 190.35250463821893

Whisper Tiny AR - Quran

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

• Loss: 1.3458
• Wer: 190.3525

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: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0001	500.0	1000	1.3154	243.7848
0.0001	1000.0	2000	1.3080	204.8237
0.0	1500.0	3000	1.3405	221.1503
0.0	2000.0	4000	1.3458	190.3525

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1