whisper-arabic-english

This model is a fine-tuned version of openai/whisper-small on farahabdou/FLEURS-AR-EN-split dataset.

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: 100
• mixed_precision_training: Native AMP

Training results

Base Model:

• BLEU: 9.80
• CHRF: 29.17

Fine-tuned Model:

• BLEU: 10.46
• CHRF: 42.97

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0