whisper-tiny_egyptian_asr

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

• Loss: 1.5403

Validation Metrices

• Wer: 85.0561
• Cer: 50.8612

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: 0.001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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
• num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
No log	1.0	47	1.1866	97.2640	54.3889
No log	2.0	94	1.0911	74.2869	41.6325
1.2016	3.0	141	1.1662	82.6445	46.5776
1.2016	4.0	188	1.2791	92.4823	50.8711
0.7033	5.0	235	1.3366	70.8441	40.7291
0.7033	6.0	282	1.5403	85.0561	50.8612

Framework versions

• Transformers 4.52.0
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.4