whisper-small-tamazight-ar

This model is a fine-tuned version of openai/whisper-small on the Tamazight Speech to Arabic Text dataset. It achieves the following results on the evaluation set:

• Loss: 2.2424
• Wer: 113.4340

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.2389	0.87	1000	2.2424	113.4340

Framework versions

• Transformers 4.39.3
• Pytorch 2.6.0+cu124
• Datasets 2.14.6
• Tokenizers 0.15.2