whisper-tiny-arabic-finetuned-v1

This model is a fine-tuned version of openai/whisper-tiny on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4792
• Wer: 45.2567
• Cer: 16.0322

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
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 50
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.5021	1.0	883	0.5576	53.2786	18.9232
0.3667	2.0	1767	0.5044	49.4192	17.9494
0.2985	3.0	2650	0.4783	47.1619	17.1923
0.2252	4.0	3532	0.4792	45.2567	16.0322

Framework versions

• Transformers 4.28.0
• Pytorch 2.6.0+cu124
• Datasets 2.18.0
• Tokenizers 0.13.3