README.md

metadata

language:
  - ar
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - ahishamm/QURANICWhisperDataset
metrics:
  - wer
model-index:
  - name: QURANIC Whisper Large V3 - 2
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: QURANICWhisperDataset
          type: ahishamm/QURANICWhisperDataset
          args: 'config: ar, split: train'
        metrics:
          - name: Wer
            type: wer
            value: 112.02681655041647

QURANIC Whisper Large V3 - 2

This model is a fine-tuned version of openai/whisper-large-v3 on the QURANICWhisperDataset dataset. It achieves the following results on the evaluation set:

• Loss: 0.1663
• Wer: 112.0268

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: 8
• 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
0.0862	2.0	1000	0.1308	162.4365
0.0489	4.0	2000	0.1305	168.4432
0.0111	6.0	3000	0.1499	193.2011
0.0013	8.0	4000	0.1663	112.0268

Framework versions

• Transformers 4.39.2
• Pytorch 2.2.0
• Datasets 2.18.0
• Tokenizers 0.15.1