Whisper Turbo - Hanif Rahman

This model is a fine-tuned version of ihanif/whisper-turbo-ar-v2 on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:

• eval_loss: 2.9315
• eval_wer_ortho: 867.0862
• eval_wer: 880.1660
• eval_runtime: 1478.3887
• eval_samples_per_second: 1.09
• eval_steps_per_second: 0.068
• epoch: 1.8779
• step: 400

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 500
• training_steps: 2000
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.46.1
• Pytorch 2.5.0+cu121
• Datasets 3.0.2
• Tokenizers 0.20.1