whisper-medium-ar-tiny

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

• Loss: 1.0894
• Wer: 36.7470

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.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: 10
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.3747	2.2222	20	1.4933	56.6265
1.1545	4.4444	40	1.2450	57.2289
1.0411	6.6667	60	1.1504	38.5542
0.99	8.8889	80	1.1043	38.5542
0.9489	11.1111	100	1.0894	36.7470

Framework versions

• Transformers 4.50.0
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1