xbilek25/whisper-small-cv_train-3600

Whisper Small Hi - Sanchit Gandhi

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

• Loss: 0.5294
• Wer: 17.4180

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 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: 360
• training_steps: 3600
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.061	2.0417	600	0.3876	16.0348
0.007	5.0208	1200	0.4369	16.2398
0.0012	7.0625	1800	0.4970	17.6742
0.0007	10.0417	2400	0.5139	17.5717
0.0005	13.0208	3000	0.5253	17.3668
0.0005	15.0625	3600	0.5294	17.4180

Framework versions

• Transformers 4.51.0
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1