Yassinevic/whisper-tiny-ar

Whisper Tiny AR - Yassinevic

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

• Loss: 0.0000
• Wer: 0.0

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: 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: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0015	9.0909	1000	0.0001	0.0
0.0001	18.1818	2000	0.0000	0.0
0.0001	27.2727	3000	0.0000	0.0
0.0	36.3636	4000	0.0000	0.0
0.0	45.4545	5000	0.0000	0.0

Framework versions

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