WhisperTiny15hCommonVoice0.0Augment

This model is a fine-tuned version of OpenAI/whisper-tiny on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.9043
• Wer: 0.4284

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: 500
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.8215	0.8929	500	1.0601	0.4950
0.6933	1.7857	1000	0.9860	0.4655
0.5697	2.6786	1500	0.9576	0.4548
0.4809	3.5714	2000	0.9356	0.4498
0.4283	4.4643	2500	0.9220	0.4484
0.3732	5.3571	3000	0.9059	0.4624
0.3367	6.25	3500	0.9070	0.4315
0.3168	7.1429	4000	0.9012	0.4284
0.311	8.0357	4500	0.8957	0.4299
0.2897	8.9286	5000	0.9043	0.4284

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0
• Datasets 3.3.0
• Tokenizers 0.21.0