leliw/whisper-tiny-pl

Whisper Tiny PL

This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 17 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6714
• Wer Ortho: 75.9211
• Wer: 66.7013

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
• 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: constant_with_warmup
• lr_scheduler_warmup_steps: 50
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.5684	0.7716	500	0.7197	103.1812	76.3039
0.4006	1.5432	1000	0.6714	79.3973	64.9667
0.2894	2.3148	1500	0.6739	78.6396	65.9231
0.2095	3.0864	2000	0.6714	75.9211	66.7013

Framework versions

• Transformers 4.50.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1