whisper-medium-en-cv-8.0

This model is a fine-tuned version of openai/whisper-medium.en on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.7352
• Wer: 22.9026

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: 48
• eval_batch_size: 32
• 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: 375
• training_steps: 2250
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0	0	2.0556	32.3025
0.5507	0.1667	375	0.7920	25.9032
0.3861	0.3333	750	0.7215	24.6479
0.205	1.1667	1125	0.7130	22.8108
0.1431	1.3333	1500	0.7193	23.8212
0.0802	2.1667	1875	0.7302	23.5150
0.0626	2.3333	2250	0.7352	22.9026

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1